<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[zkPass - Medium]]></title>
        <description><![CDATA[zkTLS Oracle Protocol for verifiable web private data. - Medium]]></description>
        <link>https://medium.com/zkpass?source=rss----1080d7b7c688---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>zkPass - Medium</title>
            <link>https://medium.com/zkpass?source=rss----1080d7b7c688---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 17:42:17 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/zkpass" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Trust Before Execution: Introducing zkPass.AI]]></title>
            <link>https://medium.com/zkpass/trust-before-execution-introducing-zkpass-ai-7e53074925c7?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/7e53074925c7</guid>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Wed, 05 Aug 2026 02:49:14 GMT</pubDate>
            <atom:updated>2026-08-05T02:49:13.141Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ik20cJvxFm1tQpcOKO8kAA.png" /></figure><p><em>AI agents can already discover services, invoke tools, coordinate work, and move value. What they still lack is a reliable way to decide whom to trust before execution.</em></p><p>The infrastructure for an open agent economy is coming into place.</p><p>Protocols such as MCP and A2A allow agents to connect with tools, services, and one another. Emerging identity standards give agents persistent, machine-readable identities. Payment protocols make it possible for software to purchase services autonomously and settle transactions over the internet.</p><p>Together, these systems answer three important questions:</p><ul><li>Where can an agent be found?</li><li>How can another system interact with it?</li><li>How can it be paid?</li></ul><p>But they do not answer the question that matters immediately before execution:</p><blockquote><em>Should this agent be trusted with the task?</em></blockquote><p>An agent may publish a valid identity, expose a well-formed manifest, and claim support for dozens of capabilities. None of that proves that its endpoint is currently available, its tools actually work, its security controls are adequate, or its performance remains consistent over time.</p><p>Discovery without verification produces a directory. Payment without verification creates economic risk.</p><p>zkPass.ai is building the missing layer between the two: a continuously updated trust infrastructure for the open agent economy.</p><h3>The verification gap</h3><p>Publishing an agent claim is inexpensive.</p><p>A developer can register an identity, deploy an endpoint, list supported capabilities, and describe the service as reliable, secure, or production-ready. A registry can record those claims, but it cannot independently guarantee that they remain true.</p><p>Verification is fundamentally different.</p><p>To verify an agent, a system must connect to it, inspect its public metadata, test protocol behavior, invoke declared tools, evaluate responses, record failures, and repeat those checks as the agent changes.</p><p>That last point is essential because agents are not static software artifacts. Their behavior can shift when any underlying component changes:</p><ul><li>the foundation model;</li><li>the system prompt;</li><li>a tool or MCP server;</li><li>a third-party API;</li><li>a permission policy;</li><li>the hosting environment;</li><li>a protocol implementation;</li><li>or the agent’s own orchestration logic.</li></ul><p>An audit performed months ago may say little about the agent running today.</p><p>Trust, therefore, cannot be treated as a permanent credential granted after a one-time review. It must be maintained as a living record supported by fresh evidence.</p><h3>What zkPass.AI is building</h3><p>zkPass.ai is an Agent Trust Network designed to produce, organize, and distribute verifiable information about agents.</p><p>At a high level, the network connects three functions:</p><ol><li><strong>Auditors produce evidence</strong> by testing agents and recording observable behavior.</li><li><strong>Trust records accumulate history</strong> across audits, tasks, changes, and risk events.</li><li><strong>Applications use that evidence</strong> to make discovery, routing, and payment decisions.</li></ol><p>The result is not simply another marketplace or leaderboard. It is a machine-readable trust layer that can be queried before an agent is given work.</p><p>The system is organized around three core components: the Auditor, the Agent Trust Record, and the Trust Graph.</p><h3>Auditor: turning claims into observable evidence</h3><p>The Auditor is the evidence-production layer of zkPass.AI.</p><p>Given an agent endpoint or identity, an Auditor can inspect the public surfaces through which the agent presents itself to the network. Depending on the agent and protocol, an audit may cover:</p><ul><li>endpoint availability;</li><li>Agent Card and metadata quality;</li><li>A2A, MCP, OASF, or related protocol behavior;</li><li>declared capabilities;</li><li>reachable tools;</li><li>request and response handling;</li><li>live task completion;</li><li>prompt-injection resistance;</li><li>unsafe or unexpected behavior;</li><li>dependency and model changes;</li><li>and machine-readable execution evidence.</li></ul><p>The objective is not to determine whether an agent sounds convincing. It is to compare what the agent claims with what it demonstrably does.</p><p>A useful audit report should therefore contain more than a score. It should expose the underlying facts:</p><ul><li>What was tested?</li><li>When was it tested?</li><li>Which capabilities were reachable?</li><li>Which requests succeeded or failed?</li><li>What warnings were observed?</li><li>Which Auditor produced the result?</li><li>Can another system verify or reproduce the evidence?</li></ul><p>This structure allows reports to be consumed by humans, marketplaces, registries, and automated routing systems.</p><h3>Agent Trust Record: a persistent history of verified behavior</h3><p>A single audit is useful, but it is not sufficient.</p><p>zkPass.ai organizes evidence around a persistent Agent Trust Record. This record connects the agent’s identity with its capability evidence, audit history, observed outcomes, dependencies, and risk events.</p><p>As the agent evolves, new evidence is added to the same record.</p><p>A model update may trigger a new audit. A failed endpoint check may reduce confidence in current availability. A completed task may contribute an observed outcome. A security warning may remain visible until it is resolved and verified.</p><p>The record therefore answers not only whether an agent passed a particular test, but also how its behavior has changed over time.</p><p>This distinction matters because two agents with similar summary scores may have very different trust profiles.</p><p>One may have been verified by several independent Auditors within the past week and maintained stable behavior for months. Another may rely on a single outdated audit performed before a major model change.</p><p>A scalar score cannot express that difference on its own. The underlying record can.</p><h3>Trust Graph: from isolated reports to network intelligence</h3><p>As more agents and Auditors participate, their individual records begin to form a Trust Graph.</p><p>The graph connects:</p><ul><li>agents;</li><li>creators and operators;</li><li>declared capabilities;</li><li>tested capabilities;</li><li>models and checkpoints;</li><li>tools and external dependencies;</li><li>Auditor nodes;</li><li>audit histories;</li><li>observed task outcomes;</li><li>and risk events.</li></ul><p>This creates a richer decision surface than a conventional directory.</p><p>An application could ask:</p><ul><li>Which agents have maintained greater than 99% availability over recent verification periods?</li><li>Which agents support a required capability and have been audited within the past seven days?</li><li>How many independent Auditors confirmed the same tool behavior?</li><li>Did the agent change its model or dependencies after its latest audit?</li><li>Do several agents share the same failing upstream service?</li><li>Which agent best satisfies the risk, latency, and capability requirements of this specific task?</li></ul><p>Over time, the Trust Graph can become more valuable than any individual audit. Testing software can be replicated, but a long-running history of independent observations, dependency relationships, changes, and outcomes cannot be recreated instantly.</p><h3>Trust cannot be reduced to one universal score</h3><p>A reputation score can be a useful summary. It should not be mistaken for the trust system itself.</p><p>Different tasks require different evidence.</p><p>A financial agent may need to demonstrate restricted permissions, deterministic transaction policies, safe contract interactions, and a history of controlled losses.</p><p>A research agent may be evaluated on source quality, citation accuracy, evidence ranking, and factual consistency.</p><p>A data agent may be judged primarily on availability, freshness, throughput, and latency.</p><p>Collapsing those dimensions into a single universal ranking would conceal the information that matters most for each task.</p><p>zkPass.ai therefore treats scores and grades as navigation aids rather than substitutes for evidence. The system’s primary product is the structured trust record beneath the summary.</p><p>This also means that “the best agent” is rarely a meaningful category. The more useful question is:</p><blockquote><em>Which agent is sufficiently verified for this task, under these constraints, at this moment?</em></blockquote><h3>From trust data to pre-execution routing</h3><p>Trust evidence becomes economically valuable when it enters the routing process.</p><p>Consider an enterprise system choosing between two agents for a funds-related operation.</p><p>Agent A is less expensive, but its most recent audit is three months old. Several declared tools no longer respond, and the operator changed the underlying model after the last verification.</p><p>Agent B costs more, but it has been evaluated recently by multiple independent Auditors. Its required capabilities remain reachable, its permission model matches the task, and its execution history satisfies the enterprise’s reliability threshold.</p><p>Without a trust layer, the router may select Agent A based on price or self-reported capabilities.</p><p>With zkPass.AI, the application can define an explicit policy:</p><blockquote><em>Route funds-related tasks only to agents verified by at least two independent Auditors within the past seven days, with full capability coverage and no unresolved critical security warnings.</em></blockquote><p>The routing flow becomes:</p><pre>Task request<br>→ Query current trust evidence<br>→ Evaluate capability, freshness, reliability, and risk<br>→ Select an eligible agent<br>→ Execute the task<br>→ Settle payment<br>→ Add the observed outcome to the trust record</pre><p>The important shift is that verification no longer exists only as a report reviewed after the fact. It becomes infrastructure that machines can use before execution.</p><p>This is the role of a trust clearing layer.</p><p>Payment clearing determines whether value was transferred correctly. Trust clearing determines whether a service should receive the task and payment in the first place.</p><h3>Accountability belongs with the Auditor</h3><p>A trust network also needs an economic model that assigns responsibility to the correct participant.</p><p>The agent is the object being observed. It does not control whether an Auditor runs the required tests, records the evidence accurately, or submits a valid report.</p><p>The Auditor controls those actions.</p><p>For that reason, zkPass.ai places economic responsibility on Auditor nodes rather than allowing capital to purchase an agent’s reputation.</p><p>Delegated ZKP expands an Auditor’s <strong>Trust Capacity</strong>: the upper limit of recurring verification responsibility that the node can assume. More delegated capital represents a larger economic commitment to the accuracy and availability of the node’s work.</p><p>Capital alone, however, does not create verification capacity or guarantee rewards.</p><p>An Auditor’s effective contribution must also reflect factors such as:</p><ul><li>completed verification work;</li><li>historical accuracy;</li><li>evidence freshness;</li><li>task completion rate;</li><li>productive capacity utilization;</li><li>and independence from other Auditors.</li></ul><p>Capital determines how much responsibility a node may accept. Verifiable contribution determines whether it deserves additional work and rewards.</p><p>This model also protects the integrity of agent evaluations. An operator may pay for faster, deeper, or more frequent testing, but it cannot pay for a higher score or a favorable conclusion.</p><h3>A public layer and an open audit market</h3><p>Not every verification request has the same purpose.</p><p>The network can support two complementary forms of audit supply.</p><p>The <strong>public verification layer</strong> maintains baseline evidence for the broader ecosystem. It can monitor endpoint health, protocol compatibility, capability availability, and recurring changes across commonly used agents.</p><p>The <strong>open audit market</strong> supports specialized demand. An enterprise may require more frequent verification, deeper security testing, a custom service-level agreement, or a multi-Auditor review for a high-value agent.</p><p>Separating these layers preserves a shared baseline while allowing advanced requirements to be funded directly.</p><p>Public trust data remains a common good. Specialized assurance becomes a commercial service. Payment can increase the depth or speed of verification, but not determine its outcome.</p><h3>A pragmatic path toward cryptographically verifiable audits</h3><p>The long-term objective is not merely to trust that an Auditor performed its work correctly. It is to make increasingly large parts of that work independently verifiable.</p><p>The network can evolve in stages.</p><p>The initial system can rely on signed, structured, and reproducible evidence. A signature proves which Auditor published a report and whether the submitted result was modified.</p><p>That does not yet prove that every underlying test was executed exactly as specified.</p><p>zkPass technology provides a path toward stronger guarantees:</p><ul><li><strong>zkTLS</strong> can prove that a real network interaction occurred between an Auditor and an agent endpoint without exposing sensitive session data.</li><li><strong>Trusted execution environments</strong> can attest that an approved, unmodified audit suite ran inside protected hardware.</li><li><strong>Zero-knowledge proofs</strong> can verify selected audit computations without revealing confidential inputs.</li><li><strong>Cryptographic execution commitments</strong> can make submitted evidence easier for independent indexers and applications to verify.</li></ul><p>As these mechanisms mature, the network can move from signed assertions toward verifiable execution.</p><p>The goal is not to introduce cryptography for its own sake. It is to reduce the amount of trust that applications must place in any individual Auditor.</p><h3>A phased network, not an overnight claim</h3><p>Building an open trust network requires operational evidence before full decentralization.</p><p>In its initial phase, zkPass.ai can prioritize a controlled set of qualified Auditor operators, reproducible audit methods, transparent off-chain calculations, and public evidence formats. Complex automated slashing should not be introduced before the network can distinguish Auditor misconduct from failures caused by the agent or its dependencies.</p><p>This phased approach keeps the system honest about what has been verified.</p><p>The progression is deliberate:</p><ol><li>coordinate public audits;</li><li>establish persistent Agent Trust Records;</li><li>measure Auditor performance;</li><li>build the Trust Graph;</li><li>integrate evidence into production routing;</li><li>open specialized audit markets;</li><li>strengthen the network with cryptographically verifiable execution.</li></ol><p>Each stage should be supported by real operational data before the next layer of economic automation is introduced.</p><h3>Who benefits from zkPass.AI?</h3><p>The network serves several participants without requiring each of them to operate the same infrastructure.</p><p><strong>Agent developers</strong> can attach credible, continuously updated evidence to their services instead of relying entirely on marketing claims.</p><p><strong>Marketplaces and registries</strong> can display capability coverage, audit freshness, reliability, and risk information alongside agent profiles.</p><p><strong>Routers and multi-agent systems</strong> can select services according to task-specific policies rather than generic rankings.</p><p><strong>Enterprises</strong> can define deterministic admission requirements without building a complete internal agent-testing stack.</p><p><strong>Auditors</strong> can contribute verification infrastructure, establish a measurable performance history, and earn responsibility through accurate work.</p><p><strong>Delegators</strong> can support accountable audit capacity rather than speculating on the reputation of individual agents.</p><p>Applications consuming trust evidence do not need to make ZKP ownership part of the end-user experience. The token’s primary role is to coordinate and secure the supply side of the verification network.</p><h3>Trust as infrastructure</h3><p>The open agent economy will not scale through connectivity and payments alone.</p><p>Agents need a way to evaluate unfamiliar services before they grant access, assign work, or release funds. Marketplaces need evidence that goes beyond self-reported profiles. Enterprises need machine-readable policies that can be enforced without manual review at every step.</p><p>zkPass.ai is building the infrastructure for those decisions.</p><p>The Auditor converts claims into observable evidence. The Agent Trust Record preserves that evidence over time. The Trust Graph turns isolated observations into network intelligence. Routing systems use that intelligence before execution. Delegated economic responsibility holds the producers of verification data accountable.</p><p>AI agents are learning how to discover services, coordinate work, and transact autonomously.</p><p>The next step is to give them a reliable way to decide whom to trust.</p><blockquote><strong><em>zkPass.AI is building the trust layer for the open agent economy, so that agents can verify before they route, execute, and pay.</em></strong></blockquote><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7e53074925c7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/trust-before-execution-introducing-zkpass-ai-7e53074925c7">Trust Before Execution: Introducing zkPass.AI</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[One Tree to Rule Them All: How GGM-Tree Optimization Makes VOLE-in-the-Head Faster]]></title>
            <link>https://medium.com/zkpass/one-tree-to-rule-them-all-how-ggm-tree-optimization-makes-vole-in-the-head-faster-cbcfd32f0049?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/cbcfd32f0049</guid>
            <category><![CDATA[updates]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Mon, 22 Jun 2026 16:03:52 GMT</pubDate>
            <atom:updated>2026-06-22T16:03:52.272Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*P2ndgcLV18y0oHuUi11cdw.png" /></figure><blockquote><em>TL;DR — zkPass turns private HTTPS data into publicly verifiable proofs using VOLE-in-the-Head (VOLEitH). The 2024 paper One Tree to Rule Them All shows that much of VOLEitH’s overhead comes from how GGM trees are built and opened. Merging τ trees into one interleaved tree, plus a proof-of-work trade-off, can shrink proofs by 10–40% without weakening security — exactly the class of optimization that can make zkPass faster and cheaper at scale.</em></blockquote><h3>Why the Verifiable Internet Needs zkPass</h3><p>TLS keeps your bank balance and your exchange KYC status private — but also unverifiable to anyone else. A screenshot proves nothing cryptographically.</p><p>zkPass closes that gap with zkTLS: users prove statements like <em>“my balance exceeds $10,000”</em> without ever exposing the raw response. Proofs run locally in the browser, then verify on-chain or inside an app.</p><p>The engine behind this is VOLE-in-the-Head (VOLEitH) — a way to make VOLE-based zero-knowledge proofs publicly verifiable, without trusted setup, at a cost a browser can actually bear. SNARKs struggle here: prover cost for TLS-sized responses is often prohibitive on the client.</p><p>So the question is not <em>whether</em> VOLEitH works. It’s how to make it smaller and faster without touching security — which is exactly what the “One Tree” paper answers.</p><h3>zkPass in Brief</h3><p>zkPass sits between Web2 privacy and Web3 verifiability. Four parties take part: the user (prover), a zkPass node, the HTTPS server, and the verifier.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E09YWRb7fb9mcKgawQ3xtA.png" /></figure><p>Most sessions use Proxy Mode: the node acts as a TLS proxy, and heavy cryptography runs only at proof time. A rarer MPC Mode handles IP-restricted servers with joint garbled-circuit TLS (~32 MB for a typical session).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*beN5mHPevlIrTtsGTVk7Ng.png" /></figure><p>For Proxy Mode — the default path — proof generation time and proof size are the UX bottleneck. That is exactly where this paper applies.</p><h3>Why VOLEitH, not SNARKs?</h3><p>zkPass could have built on SNARKs. It chose VOLE-in-the-Head instead, and the trade-offs explain why:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RPPbrTncOORJpf1Mce74Ig.png" /></figure><p>Standard VOLE-ZK is designated-verifier: only the party holding secret VOLE key material can check the proof. VOLEitH (Baum et al., CRYPTO 2023) fixes that — a publicly verifiable NIZK with roughly 2× smaller proofs than MPC-in-the-Head, under the same assumptions, with similar or better runtime.</p><h3>How VOLEitH Works: Three Phases</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7-Z7iwcWHTuDpSSz-3nQBQ.png" /></figure><p>Phase 1 builds GGM vector commitments. Phase 2 runs QuickSilver over the schema circuit (e.g., proving balance &gt; 10000 in a JSON response). Phase 3 opens everything for public verification. The paper optimizes Phases 1 and 3 — the GGM-tree cost — without touching circuit logic.</p><h3>Where the Cost Hides: GGM Trees</h3><p>VOLEitH builds GGM trees (Goldreich–Goldwasser–Micali): expand a random seed with a PRG into a binary tree, commit to the leaves, then open all-but-one leaf per vector by revealing internal nodes along the paths.</p><p>Before optimization, VOLEitH maintained a forest of τ separate trees:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nXR0VS0rNg6XobFvleqycg.png" /></figure><p>These openings are not a footnote. In MPC-in-the-Head schemes like Picnic, they dominate proof size. VOLEitH already cuts size roughly in half versus MPCitH — but GGM openings remain the largest piece left to optimize.</p><h3>One Tree, Three Optimizations</h3><p>Baum, Beullens, Mukherjee, Orsini, Ramacher, Rechberger, Roy, and Scholl (ASIACRYPT 2024) stack three improvements on top of one another.</p><h4>Batch All-But-One Vector Commitments (BAVC)</h4><p>Instead of opening each GGM tree independently, the authors formalize BAVC — batch vector commitments opened together under a single challenge. Independent openings pay the full log(N) · λ cost per vector; structured batching shares opening information across vectors.</p><h4>One Interleaved Tree</h4><p>The central move: merge τ trees into one tree with interleaved leaves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*flNdqNM7942ayAkyMvIcGQ.png" /></figure><p>Why interleave? Contiguous placement would spread hidden leaves far apart, forcing more internal nodes open. Interleaving keeps them adjacent. An opening threshold T_open caps how many internal nodes may be revealed; if it&#39;s exceeded, the prover retries — so proof sizes stay predictable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Xck_N0D1d-fh7GlZYI5B8g.png" /></figure><h3>4.3 Grinding and Proof-of-Work</h3><p>Merging trees introduces aborts: some challenges exceed T_open. The prover retries with an incrementing counter until success. Each retry is one hash evaluation — effectively proof of work. An attacker forging a proof must do the same.</p><p>Grinding goes further: give up w bits of challenge entropy and recover security through ~2^w hash iterations. Counterintuitively, this can speed up signing, because smaller vector commitments save cost elsewhere.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rsfbj7JolIKjMS99RQswPA.png" /></figure><p>On FAEST-class parameters, BAVC plus grinding yields roughly 8–10% smaller proofs and ~25% faster signing for size-optimized variants. Combined with AES-EM and Rain/MQ one-way functions, reductions reach 40% or more — down to sub-3 KB proofs, the smallest VOLEitH/MPCitH proofs measured to date.</p><h4>OWF Choice and QuickSilver Tuning</h4><p>VOLEitH proves knowledge of a preimage to a one-way function (OWF). The OWF choice sets the nonlinear constraint count — the main driver of QuickSilver cost, and the same mechanism that governs zkPass schema circuits.</p><p>FAEST-d7 proves AES with degree-7 constraints over GF(2) instead of degree-2 over GF(²⁸): halving witness size at 128-bit security, using the full AES key space, and trimming signatures by up to ~5%. Directly relevant to zkPass, which uses QuickSilver for circuit satisfaction.</p><h3>What This Means for zkPass</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xTEmjbHVtAUVZS6StqqFIw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E8HPTNQs6q2n3gMyQlRj9w.png" /></figure><p>zkPass has already optimized heavily elsewhere: Silent OT and stacked garbled circuits for MPC Mode (&gt;3× efficiency), hybrid Proxy/MPC switching, and 300+ reusable schemas in Circuit Factory. The VOLEitH commitment layer — this paper’s subject — is the natural next frontier for Proxy Mode.</p><p>At zkPass’s current scale (200M+ verifiable computations processed, 80+ integration partners), even a 10–20% proof-size reduction compounds into real savings: lower on-chain verification cost, less bandwidth, faster attestation in the browser.</p><p>If zkPass’s VOLEitH layer today resembles FAEST-class parameters, adopting FAESTER-style BAVC could plausibly deliver 8–15% smaller proofs, 20–30% faster proving in size-optimized mode, and proportional verification savings. <em>(Actual gains depend on schema complexity and implementation choices — these are projections, not published zkPass benchmarks.)</em></p><h3>Security: Smaller Does Not Mean Weaker</h3><p>Three concerns, three answers:</p><ul><li>Does abort/retry weaken security? No. Each retry feeds an incrementing counter into the hash. The FAEST security proof still applies: for each hash query, at most two challenges can lead to forgery, and retries are unpredictable to the adversary.</li><li>Does grinding (giving up w bits) weaken security? Not at the scheme level. Lost challenge entropy is compensated by ~2^w hash iterations, and forgers must match that work.</li><li>Does merging trees break binding or hiding? No. The paper defines new BAVC security notions (extractable binding, n-hiding real-or-random) that plug into the existing FAEST compilation with essentially the same bounds.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dIdlQobaiaxepUYTEMqnsw.png" /></figure><h3>Conclusion</h3><p><em>One Tree to Rule Them All</em> shows that a large share of VOLEitH’s cost lives not in circuit logic but in how GGM trees are built and opened. Merging τ trees into one interleaved tree, treating rejection as proof-of-work, and tuning QuickSilver for higher-degree constraints deliver meaningful size and runtime gains — up to 40%+ with specialized OWFs — without new assumptions.</p><p>For zkPass, these optimizations hit the proof system powering zkTLS. Proxy Mode lives or dies on client-side efficiency, and BAVC-style batching plus constraint tuning offer a provably secure path to faster browser proofs and cheaper on-chain verification.</p><p>The verifiable internet doesn’t need a revolution in zero-knowledge theory. It needs the careful engineering that turns VOLE-in-the-Head from a cryptographic framework into infrastructure you can run in a browser tab.</p><p>References</p><ol><li>Baum et al., <em>One Tree to Rule Them All: Optimizing GGM Trees and OWFs for Post-Quantum Signatures</em>, ASIACRYPT 2024. DOI: 10.1007/978–981–96–0875–1_15</li><li>Baum et al., <em>FAEST</em> / VOLE-in-the-Head, CRYPTO 2023</li><li><a href="https://docs.zkpass.org/overview/technical-overview">zkPass Technical Overview</a></li><li><a href="https://paper.zkpass.org/zkPass_WP2025.pdf">zkPass Whitepaper (2025)</a></li><li>Boyle et al., VOLE-ZK / QuickSilver, EUROCRYPT 2022</li></ol><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cbcfd32f0049" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/one-tree-to-rule-them-all-how-ggm-tree-optimization-makes-vole-in-the-head-faster-cbcfd32f0049">One Tree to Rule Them All: How GGM-Tree Optimization Makes VOLE-in-the-Head Faster</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Proof of Reserve for RWAs]]></title>
            <link>https://medium.com/zkpass/proof-of-reserve-for-rwas-594315c03575?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/594315c03575</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[updates]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Sat, 13 Jun 2026 15:23:59 GMT</pubDate>
            <atom:updated>2026-06-13T15:23:59.429Z</atom:updated>
            <content:encoded><![CDATA[<p><em>From the SpaceX pre-IPO token rush to a reserve check anyone can run.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NHhWU0cFyufRZH0ZSRirZg.png" /></figure><h4>The “Trust Gap” in Tokenized Assets</h4><p>In 2026 a wave of SpaceX-linked tokens hit the market. Primary-market exposure that used to sit behind institutional gates and accredited-investor checks started trading as on-chain assets, open to anyone with a wallet. Equities, funds, and bonds are moving on-chain faster than in any prior cycle. The pitch is clean: take the value of a real asset, give it the liquidity of a blockchain.</p><p>Scale exposes the crack. The moment you tokenize a physical asset, the network faces a question it cannot answer on its own. Does the thing actually exist?</p><p>Bitcoin never has this problem. Any node can read the supply and trace every coin, because verification is simply what the chain does. An RWA breaks that loop. The asset sits in a brokerage account, the ledger sits on-chain, and nothing cryptographically connects the two. You can pull the token’s total supply and trade it all day, yet you cannot reach into the broker’s system and confirm the shares are really there. The token says it is backed by real shares. The holdings data stays locked inside an institution.</p><p>With no verifiable link between them, the holder is left trusting whatever the issuer says. The whole RWA stack balances on a single unproven assumption: the off-chain asset is real.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tzfojKmNW8kDyxO-wI51_g.png" /></figure><h4><strong>Where proof of reserve falls short</strong></h4><p>The RWA market does not need more asset types. It needs a way to prove the assets are real in a form a machine can check.</p><p>Traditional finance already verifies assets through custody statements, audit reports, and disclosures. Those mechanisms work inside legacy markets, but they all resolve to the same place. You end up trusting the auditor’s reputation rather than the asset itself.</p><p>Drop that model into a market that never closes and three gaps open up.</p><ol><li><strong>Frequency. </strong>Audits run quarterly or yearly. Between them the reserve is a black box. A months-old snapshot means little to a market that trades around the clock.</li><li><strong>Latency. </strong>From kickoff to published report takes weeks. By the time you read the conclusion, it describes a state that has already moved.</li><li><strong>Granularity. </strong>Reports grade overall health, the financially-sound verdict, not a specific asset at a specific second. You cannot point at one token and check the exact shares behind it on demand.</li></ol><p>Tooling has tried to patch this. Chainlink’s Proof of Reserve uses oracles to pull data from a custodian or auditor API on a schedule and post it on-chain, which helps update frequency. But it is a relay, not a proof. Whether the number is true still rides on trusting the data provider and the oracle nodes. Nothing at the source carries a cryptographic guarantee.</p><h4>The Verification Paradigm Shift: From Institutional Endorsement to Cryptographic Fact</h4><p>Legacy finance answers one question: who should you trust? An RWA system has to answer a different one: how does the asset prove itself? Those are two different engineering problems.</p><p>If audits cannot remove the trust assumption, the move is to prove the data at its source. That is what zkTLS does.</p><p>Take a stock token’s reserve. The old way: the platform shows you a holdings report, you see a number, and you have no way to trace where it came from. zkTLS changes the requirement. You do not trust the platform. Cryptography proves where the data originated.</p><p>When the asset controller logs into the broker, the two servers open a TLS channel, the same encryption that secures banking and government portals. zkTLS takes that live session and produces a zero-knowledge proof over it. The proof establishes four things to anyone watching:</p><ul><li>The data came from the broker’s own server, that specific machine.</li><li>It is the raw server response, with no human in the loop.</li><li>It survived transmission intact, with no tampering.</li><li>The numbers are exact: 1,000,000 AAPL shares, not roughly a million.</li></ul><p>At the same time, login state, credentials, and every unrelated position in the account stay sealed behind the proof. The reserve number goes public. The private data never leaves.</p><p>The goal is narrow and specific. Turn a claim, that this broker account held this much of this asset at this time, into a fact anyone can call and check for themselves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IHC5LThbuiA_wNAwut_R0A.png" /></figure><h4>Trust-Minimized System Design and Separation of Duties</h4><p>The architecture has one non-negotiable rule. No single participant can fake a reserve. To hold that line, the system draws two boundaries that nothing crosses.</p><p><strong>Boundary 1, data authenticity. </strong>This sits between the Prover, the proof execution layer, and the broker API. No screenshots, no static reports, no internal database dumps. The only accepted input is a live TLS session opened directly between the Prover node and the broker’s server. Authenticity comes off the wire, not off a declaration.</p><p><strong>Boundary 2, proof credibility. </strong>A proof cannot vouch for itself. It goes to the decentralized zkPass network, where nodes run the ZK computation and sign the result. That signature is non-repudiable and can be written to a smart contract as a permanent attestation anyone can reference.</p><p>Holding it together is a strict split of duties and credentials:</p><ul><li>Platform (backend). Routing, billing, state machines. It never touches broker credentials, so it physically cannot change a result.</li><li>Prover (proof execution layer). A stateless node. Broker credentials live in memory only for the seconds a proof takes, then they are gone.</li><li>Verification network. Fully independent, there to sign and nothing else.</li></ul><p>Under those constraints the issuer cannot invent holdings, the platform cannot edit results, and no Prover node can forge a proof on its own. Reliability stops depending on anyone behaving well. It rests on a setup where a single bad actor cannot move the outcome.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bvIa6vSn8V2FuieyIMYEHA.png" /></figure><h4>Four Core Questions Traditional Proof of Reserve Cannot Answer</h4><p>Any serious desk evaluating a PoR design pokes at the same weak points. Here is where this one holds.</p><p><strong>Q1. A proof is a snapshot. What stops someone moving the assets the second after they are verified?</strong></p><p>Every proof of reserve is point-in-time. That is physics, not a flaw. What changes is the cost of taking a fresh one. A traditional audit takes weeks. A zkTLS check costs almost nothing. As long as the controller leaves read-only API access open, any participant can fire a new proof whenever they want. When re-verifying is one click, the window between proofs stops being an attack surface. The logic is binary: an expired proof is invalid, and refusing a fresh check counts as failing it.</p><p><strong>Q2. What stops the same underlying asset being pledged to several token projects at once?</strong></p><p>During generation, the system pulls a unique account identifier from the broker’s response, hashes it into a Nullifier, and embeds it in the attestation. Every proof from the same broker account, across every token, is therefore linked in the on-chain registry. Because each attestation publishes the exact holding, say 1,000,000 shares, a contract can simply add them up. Total endorsed value under one Nullifier can never exceed the latest verified holding. Move the assets to a fresh account to mint a new proof, and the original account fails its next verification. This only works if every attestation lands in one shared on-chain registry, which is why chain integration is a requirement here, not a nice-to-have.</p><p><strong>Q3. PoR only covers assets. What about hidden liabilities? The FTX problem.</strong></p><p>FTX exposed the blind spot in exchange PoR. Deposits, the liabilities, were buried in an opaque internal database, so proving the asset side meant nothing. RWAs change the shape of the balance sheet. The token’s total on-chain supply is the liability, and it is transparent by construction. Anyone can pull the circulating supply from the contract and check it against the zkTLS asset-side proof. Both sides of the ratio become verifiable at once: liabilities by chain consensus, assets by cryptography.</p><p><strong>Q4. Aren’t you just moving the trust from auditors to brokers?</strong></p><p>Every verification model leans on some underlying data source. Traditional auditors mail confirmation letters to the same brokers. zkTLS does not add a trust assumption. It removes the intermediaries with both the reason and the room to manipulate: issuers, custodians, auditors. As for the broker, it is the most heavily regulated party in the chain. It holds no crypto, takes no cut of token issuance, and has nothing to gain from faking a number. And the system reads from the broker’s core production APIs, the ones serving its global client base. Forging one external proof would mean compromising the broker’s entire trading backbone, which is neither technically nor legally realistic.</p><h4>From Declaration to Cryptographic Fact: End-to-End Verification Workflow and Infrastructure</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fL9OdBVsrlBhD81BFIWtIA.png" /></figure><p>Take 1,000,000 AAPL tokens. A trustless reserve proof runs in four steps.</p><ol><li><strong>Initiate Verification:</strong> End users or smart contracts trigger a verification request against the target broker account’s AAPL holdings.</li><li><strong>Connect to True Source:</strong> The Prover node invokes a dedicated connector, establishes a live TLS encrypted session with the broker’s API, and securely extracts the holding data.</li><li><strong>Generate &amp; Verify ZKP:</strong> The session data is submitted to the decentralized zkPass network, which executes the zero-knowledge computation, outputs a deterministic result (AAPL = 1,000,000), and finalizes it with an irrefutable signature.</li><li><strong>Publish Results:</strong> The platform frontend verifies the signature and publishes the result. The attestation is committed on-chain, becoming an objective fact universally referable by any smart contract.</li></ol><p>To keep the cryptography out of the integrator’s way, zkPass wraps the infrastructure in a Portal. The issuer declares which data fields to prove, and the system handles schema setup, proof generation, on-chain verification, and publication.</p><p>Docs, API specs, and end-to-end examples are open source: <a href="https://github.com/zkPassOfficial/zkPass-Portal-Proof-of-Reserve">github.com/zkPassOfficial/zkPass-Portal-Proof-of-Reserve</a>. You integrate without touching the underlying ZK math.</p><h4>Conclusion: From Tokenization to Verification</h4><p>For a few years the RWA story has been tokenization, giving off-chain assets global digital liquidity for the first time. Real progress. But if an asset shows up on-chain while its authenticity stays trapped in an off-chain black box, the ecosystem keeps settling for slow audits and centralized trust.</p><p>Tokenization answers how an asset gets onto the chain. The harder question is how it earns the chain’s trust once it is there. So tokenization is not the finish line for RWAs. Verification is.</p><p>zkPass Verifiable Proof of Reserve does more than enable new asset types. It puts a verification layer between the real world and the network, beholden to neither. Traditional assets earned trust through a central institution’s endorsement. On-chain, assets establish facts through proofs that cannot be altered.</p><p>The RWA sector completes its real leap the day an asset’s authenticity can be checked by math instead of taken on someone’s word.</p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=594315c03575" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/proof-of-reserve-for-rwas-594315c03575">Proof of Reserve for RWAs</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Simplifying Byzantine Complexity in the zkPass Network with TEE]]></title>
            <link>https://medium.com/zkpass/simplifying-byzantine-complexity-in-the-zkpass-network-with-tee-ee2e418affc2?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/ee2e418affc2</guid>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Mon, 01 Jun 2026 03:43:01 GMT</pubDate>
            <atom:updated>2026-06-01T03:43:01.520Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NhO7nuKqgA7v2F1mQgBeKw.png" /></figure><p>For years, internet security has been discussed at two very different scales. At the product scale, teams ask practical questions like latency, fraud rate, and operating cost. At the civilization scale, people ask who controls computation, who can inspect it, and who can be silently excluded. A verifier architecture is not just a backend implementation detail; it is a choice about where trust lives.</p><p>In January 2026, Vitalik Buterin described this direction as “taking back lost ground in computing self-sovereignty” (<a href="https://x.com/VitalikButerin/status/2014454477152256415">x.com/VitalikButerin/status/2014454477152256415</a>): move more of the stack toward open, local, and user-controlled systems, and combine imperfect privacy technologies instead of waiting for one perfect primitive. That framing is useful here. A zkPass + TEE verifier is not only a way to simplify Byzantine operations. It is also part of a broader transition from institution-trusted computing to evidence-trusted computing.</p><p>zkPass is based on VOLE-based ZK and VOLE-in-the-Head technologies, allowing users to generate verifiable proofs without exposing their original private data. When zkPass proofs are applied in real products, a common engineering question appears: who guarantees that the verifier executes honestly? If the verifier is a normal server, the system often needs multiple verifiers to validate together, decide the result through a Byzantine threshold, and rely on fishermen to detect cheating.</p><p>TEE provides another engineering path: place the official verifier inside a measurable enclave image, prove the code hash through remote attestation, and let the enclave sign the verification result. The business backend and on-chain contract only consume lightweight signed credentials. In this article, TEE primarily refers to AWS Nitro Enclave as the target deployment environment.</p><p>In this model, TEE does not replace zkPass’s cryptographic security. zkPass still relies on the VOLE-based ZK / VOLE-in-the-Head proof system to complete privacy-preserving proofs. TEE solves the trustworthiness problem around the verifier execution environment and signed credentials.</p><h3>From Verifier Correctness to Computing Self-Sovereignty</h3><p>If we zoom out, this is not only a “how do we verify proofs cheaper” question. It is part of a larger redesign of digital power boundaries:</p><ul><li>users should reveal less raw data to centralized services;</li><li>applications should consume minimal claims instead of full identity payloads;</li><li>operators should prove runtime integrity rather than ask for institutional trust;</li><li>trust should be layered, with cryptography, hardware attestation, and decentralized audit each covering different failure modes.</li></ul><p>This is exactly why zkPass and TEE fit together. zkPass minimizes what must be revealed. TEE minimizes what must be trusted about execution. Signed credentials minimize what downstream systems must recompute. In the same spirit as the “local-first and sovereignty-first” software movement, the architecture pushes sensitive computation and key material closer to controlled boundaries, while exporting only verifiable artifacts.</p><h3>Traditional Path</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k19aaFAk_C2ksvLlmsKrlw.png" /><figcaption>Traditional multi-verifier path with Byzantine threshold</figcaption></figure><p>The benefit of this model is that it avoids trusting a single verifier, but the costs are also obvious:</p><ul><li>Multiple verifiers increase latency and cost.</li><li>Threshold protocols, staking, slashing, and dispute windows increase product complexity.</li><li>Fishermen need the ability to recompute or review results, and must continuously observe the system online.</li></ul><h3>TEE Honest Verifier</h3><p>The TEE model rewrites the problem as:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2ZN7WfiiZkcpNdy9TGM5Rw.png" /><figcaption>TEE honest-verifier path: attestation replaces the verifier committee</figcaption></figure><p>Honest execution by the verifier no longer mainly depends on multi-party games. Instead, it depends on:</p><ul><li>the verifier EIF image hash being an officially released version;</li><li>the Nitro NSM attestation document binding ImageSha384, PCRs, nonce, and the enclave signing public key;</li><li>the enclave private key never leaving the enclave;</li><li>the credential signature covering the statement hash, nullifier, policy version, result, and expiry.</li></ul><h3>Why a TEE Verifier Is Trustworthy</h3><p>Here, “trustworthy” does not mean blindly trusting a particular server, nor does it mean assuming the operator is honest. A TEE verifier is trustworthy because there is an externally checkable evidence chain behind the verification result:</p><ol><li>The verifier source code and dependencies are built into a deterministic version of the enclave image.</li><li>When Nitro Enclave starts, it generates measurements for the image, kernel/bootstrap, and application code.</li><li>NSM generates an attestation document that binds the measurement, nonce, and enclave public key together.</li><li>The AWS Nitro attestation certificate chain proves that this document comes from a real Nitro Enclave.</li><li>KMS only releases keys to enclaves that satisfy the measurement conditions.</li><li>The enclave issues credentials with its internal private key.</li><li>The on-chain contract, backend, or fisherman verifies the quote, measurement, public-key binding, and credential signature.</li></ol><p>In other words, the business side does not trust “the machine running the verifier”. It trusts that “a certain auditable verifier image is indeed running inside a Nitro Enclave, and only that image can obtain signing capability”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tW2IuGcTD9CvUePGW-dICA.png" /><figcaption>The externally checkable evidence chain from source code to accepted credential</figcaption></figure><p>Each piece of evidence in this chain has a clear purpose:</p><ul><li>ImageSha384 proves which EIF is running.</li><li>PCR0/PCR1/PCR2 prove that the combination of image, kernel/bootstrap, and application-level verifier has not been replaced.</li><li>nonce prevents old attestation documents from being replayed.</li><li>public_key binds the enclave&#39;s internal signing public key to this attestation.</li><li>The KMS recipient policy ensures that only enclaves matching the required measurements can unseal the signing key or data key.</li><li>The credential signature proves that the business result was issued by an attested enclave.</li></ul><p>Therefore, the “majority of nodes execute honestly” assumption in a normal verifier network is replaced by Nitro’s hardware isolation, verifiable measurements, KMS unsealing policy, and signed credential verification.</p><h3>How Nitro Enclave Fits This Model</h3><p>The key to Nitro Enclave is measurement. When building an EIF, nitro-cli build-enclave outputs ImageSha384 and a set of PCRs. The business side registers these values as trusted verifier versions, and the KMS policy can also bind to these measurements.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UJ8D87DfkHMlUBt8KBBjNQ.png" /><figcaption>End-to-end lifecycle across Builder, KMS, Enclave, Backend, and Consumer</figcaption></figure><p>More completely, Nitro Enclave’s role in this product can be split into three layers:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gPPHTpDtYCFr-KgCg2c3Mw.png" /><figcaption>Three layers: Build Time, Runtime, and Consumption</figcaption></figure><h3>Build Phase</h3><p>The team first packages the zkPass verifier, VOLE-based ZK / VOLE-in-the-Head proof verification logic, policy-checking logic, and signing module into the enclave image. After nitro-cli build-enclave generates the EIF, it produces the measurement. This measurement should be recorded in two places:</p><ul><li>release registry: the backend, contract, or fisherman knows which ImageSha384 values are official versions;</li><li>KMS key policy: only the enclave corresponding to this image can unseal the key.</li></ul><h3>Runtime Phase</h3><p>The backend cannot directly read enclave memory or directly call the enclave private key. It can only send the proof and public signals to the enclave over vsock. Inside the enclave, three things happen:</p><ol><li>Run the zkPass verifier.</li><li>Check the business policy, such as issuer, threshold, and policy version.</li><li>Sign the verification result, nullifier, expiry, and policy version.</li></ol><h3>Consumption Phase</h3><p>The on-chain or backend consumer does not need to re-execute the zkPass verifier. It only needs to verify:</p><ul><li>whether the credential signature comes from the enclave public key bound in the quote;</li><li>whether the ImageSha384 and PCRs in the quote are in the trusted list;</li><li>whether the quote is fresh and the nonce matches;</li><li>whether the credential has expired;</li><li>whether the nullifier has already been used.</li></ul><p>In the demo, internal/nitro uses Go types to simulate these Nitro concepts:</p><pre>type Image struct {<br>    Name          string<br>    Version       string<br>    EIFPath       string<br>    ImageSha384   string<br>    Measurements  map[string]string<br>    VerifierDigest string<br>}</pre><pre>type AttestationDocument struct {<br>    ModuleID   string<br>    Digest     string<br>    PCRs       map[string]string<br>    PublicKey  string<br>    Nonce      string<br>}</pre><p>PCR0 represents the EIF image measurement, PCR1 represents kernel/bootstrap, and PCR2 represents the verifier application. A real Nitro document is CBOR and signed by the AWS certificate chain. For local testability, the demo uses Ed25519 to simulate the attestation authority.</p><h3>Detailed Nitro Evidence Chain</h3><p>From proof to claim, trustworthiness is not provided by a single signature. It is produced by multiple layered binding relationships:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-5wqNVwGzWOHFViUUuVcIg.png" /><figcaption>Layered binding relationships from proof to consumer verification</figcaption></figure><p>The two most important bindings in this diagram are:</p><ul><li>NSM document -&gt; enclave public key: proves that this public key belongs to a measured enclave.</li><li>credential -&gt; enclave public key: proves that the business result was signed by the same enclave private key.</li></ul><p>If both bindings hold, and the measurement is the official zkPass verifier image, then the consumer can accept that “this result was issued after the official zkPass verifier honestly executed inside a Nitro Enclave”.</p><p>KMS further strengthens this conclusion. Even if the operator controls the EC2 host, they cannot move the private key onto the host to sign, because the private key can be designed to be:</p><ul><li>generated inside the enclave and never exported;</li><li>or encrypted and sealed by a KMS data key;</li><li>unsealed only after submitting an NSM attestation document;</li><li>released by KMS only after the policy checks ImageSha384 and PCRs.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XLfCWCKTtahgiQuDD_wNOg.png" /><figcaption>KMS unsealing: the host cannot decrypt without a valid attestation document</figcaption></figure><p>This is also why a TEE verifier can reduce the complexity of a Byzantine verifier network: it does not require multiple ordinary servers to supervise each other for every proof computation. Instead, it lets each verification result carry verifiable evidence of the execution environment.</p><p>It is important to note that a TEE verifier does not eliminate all trust assumptions. It narrows them into fewer, easier-to-audit boundaries:</p><ul><li>The system must trust AWS Nitro’s hardware isolation and attestation root.</li><li>The verifier image must be reproducibly built, or at least the publisher must be able to stably produce and register measurements.</li><li>The KMS policy, trusted image registry, and policy version governance must be correct.</li><li>The consumer must strictly check the quote, credential signature, expiry, and nullifier.</li><li>Fishermen or monitoring systems are still needed to detect old versions, misconfigurations, denial of service, and abnormal signatures.</li></ul><p>So a more precise statement is: a TEE verifier turns the game-theoretic question of “whether multiple verifiers execute honestly” into the verifiable engineering question of “whether the specified verifier program executed inside the specified TEE measurement”.</p><h3>Core Boundaries in the Model</h3><p>VerifyAndAttest in internal/tee is the main flow: it first executes the verifier and policy checks, then issues the credential.</p><pre>cred := EligibilityAttestation{<br>    ProofType:        p.Type,<br>    StatementHash:    result.StatementHash,<br>    UserWallet:       signals.UserWallet,<br>    Nullifier:        proof.Nullifier(signals, p),<br>    PolicyVersion:    e.policy.Version,<br>    Result:           status,<br>    IssuedAt:         now.Unix(),<br>    ExpiresAt:        now.Add(e.policy.CredentialTTL).Unix(),<br>    EnclavePublicKey: hex.EncodeToString(e.publicKey),<br>}<br>cred.Quote = e.attestationService.IssueNitroQuote(e.nitroImage, e.publicKey, e.policy.V…)<br>cred.Signature = hex.EncodeToString(ed25519.Sign(e.privateKey, CredentialSigningBytes(c…)</pre><p>The quote now contains a Nitro-style attestation document internally. The business layer still only sees a generic quote, but the security semantics are already close to Nitro:</p><pre>document := nitro.NewAttestationDocument(<br>    &quot;i-local.nitro-enclave&quot;,<br>    image,<br>    signingPublicKey,<br>    policyVersion,<br>    nonce,<br>    now,<br>)</pre><p>internal/chain simulates an on-chain contract. It does not re-execute the proof verifier. It only checks whether the credential comes from a trusted enclave and prevents replay.</p><pre>if err := tee.VerifyCredentialSignature(cred); err != nil {<br>    return Receipt{}, err<br>}<br>if err := c.attestationService.VerifyQuote(cred.Quote, attestation.VerifyOptions{<br>    TrustedImageHashes: c.trustedImages,<br>    SigningPublicKey:   signingKey,<br>    MinPolicyVersion:   c.minPolicyVersion,<br>    Now:                now,<br>}); err != nil {<br>    return Receipt{}, err<br>}</pre><p>Quote verification further checks the Nitro document:</p><pre>nitro.VerifyDocument(q.Document, authorityKey, nitro.VerifyOptions{<br>    ExpectedImageSha384: q.ImageHash,<br>    ExpectedPCRs:        q.Image.Measurements,<br>    ExpectedPublicKey:   signingKey,<br>    ExpectedNonce:       q.Nonce,<br>})</pre><p>If the system needs to protect the signing key or verifier configuration inside the enclave, KMS can be configured to unseal the data key only for enclaves matching the required measurement. The demo’s KMSRecipientConditions correspond to the condition keys in a real KMS policy:</p><pre>{<br>  &quot;kms:RecipientAttestation:ImageSha384&quot;: &quot;&lt;image-sha384&gt;&quot;,<br>  &quot;kms:RecipientAttestation:PCR0&quot;: &quot;&lt;pcr0&gt;&quot;,<br>  &quot;kms:RecipientAttestation:PCR1&quot;: &quot;&lt;pcr1&gt;&quot;,<br>  &quot;kms:RecipientAttestation:PCR2&quot;: &quot;&lt;pcr2&gt;&quot;<br>}</pre><h3>How the Fisherman’s Role Changes</h3><p>With TEE, fishermen do not necessarily disappear, but their responsibility changes from “recomputing proofs and challenging verifiers” to “supervising trusted execution boundaries”:</p><ul><li>whether ImageSha384 and PCRs are official versions;</li><li>whether the NSM attestation document was issued by a trusted Nitro attestation chain;</li><li>whether the quote binds the credential’s signing key;</li><li>whether the policy version is too old;</li><li>whether the credential has expired;</li><li>whether the same nullifier has conflicting results.</li></ul><p>This makes fishermen more like security observers between on-chain and off-chain systems, rather than duplicate verifiers for every proof.</p><h3>Composable Sovereignty Stack (Inspired by Vitalik’s 2026 Framing)</h3><p>Vitalik’s post also highlights a practical engineering attitude for privacy-preserving systems: combine multiple imperfect tools into a better end-to-end posture. Applied to zkPass verifier infrastructure, this can be read as a four-layer strategy:</p><ol><li>Local-first pre-processing: reduce sensitive data at the edge before remote calls.</li><li>Zero-knowledge proving: prove eligibility without exposing underlying records.</li><li>TEE-based honest execution: attest verifier runtime and bind signatures to measured code.</li><li>On-chain or federated audit: enforce replay protection, version policy, and transparent accountability.</li></ol><p>The same mindset appears in Vitalik’s concrete software choices in that thread: replacing default centralized tools with open and privacy-preserving alternatives such as Signal, OpenStreetMap, and ProtonMail. The point is not ideological purity or a one-step migration. The point is to reduce unnecessary trust concentration surface by surface, while keeping systems usable.</p><p>For AI-heavy workloads, this composition can evolve further: local small models for filtering and redaction, TEE for confidential execution when off-device compute is needed, and privacy-preserving payment and authorization flows around each query. Long-term, more efficient FHE may absorb parts of this stack. Until then, the point is not perfection, but layered risk reduction with explicit trust boundaries.</p><h3>Simulation and Production Differences</h3><p>The demo’s MockVOLEVerifier only uses deterministic digests to simulate the input/output boundary of a real zkPass verifier. It does not implement real VOLE-based ZK or VOLE-in-the-Head cryptography. attestation.Service and internal/nitro.AttestationDocument are also only local attestation authority simulators, and do not represent a real Nitro Enclave CBOR document or AWS certificate chain.</p><p>For production, the following components need to be replaced:</p><ul><li>a real zkPass verifier;</li><li>Nitro Enclave NSM attestation document parsing and AWS certificate chain verification;</li><li>signing key generation inside the enclave, with unsealing through KMS recipient attestation;</li><li>on-chain verification of quote-derived trust roots or maintenance of a trusted enclave public-key set;</li><li>release, revocation, upgrade, and fisherman alerting processes.</li></ul><p>The value of this model is that it makes the system boundaries explicit: the zkPass proof verifier, TEE attestation, backend, chain consumer, and fisherman each take responsibility only for their own role.</p><p>At a larger scale, this is a template for rebuilding trustworthy internet infrastructure: do as much as possible with local control, prove what cannot be local with cryptography, attest what must run remotely, and keep public accountability where coordination is required. The verifier is only one component, but it shows how systems can move from “please trust us” to “here is the evidence chain”.</p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ee2e418affc2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/simplifying-byzantine-complexity-in-the-zkpass-network-with-tee-ee2e418affc2">Simplifying Byzantine Complexity in the zkPass Network with TEE</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Verifiable Reputation Score (VRS) — An Open Framework by zkPass]]></title>
            <link>https://medium.com/zkpass/verifiable-reputation-score-vrs-an-open-framework-by-zkpass-31120a3a9329?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/31120a3a9329</guid>
            <category><![CDATA[updates]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Sun, 26 Oct 2025 10:53:29 GMT</pubDate>
            <atom:updated>2025-10-26T11:06:14.502Z</atom:updated>
            <content:encoded><![CDATA[<h3>Verifiable Reputation Score (VRS) - An Open Framework by zkPass</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_rrkxB8uYJ-q-p-v-rZp2Q.png" /></figure><p>The internet was built for communication, not for credibility.<br>Every profile, rating, and credential we see online is based on visibility, not verifiability. Screenshots can be faked, metrics can be inflated, and identity can be imitated. The digital economy runs on information that is easy to share but impossible to prove.</p><p>This problem becomes more severe as the internet merges with financial systems. DeFi, airdrops, credit protocols, and prediction markets increasingly depend on who participates, not just what happens on-chain. Yet most reputation signals still come from Web2 platforms where private data is trapped behind centralized APIs. The result is an asymmetry between <em>what is visible</em> and <em>what is verifiable</em>.</p><p>zkPass exists to close that gap.<br>Powered by the <strong>zkTLS protocol</strong>, zkPass enables users to prove facts about their Web2 accounts, credentials, and behaviors without exposing any underlying data. It transforms ordinary interactions - like finishing a Duolingo lesson, holding a verified LinkedIn account, or completing a KYC process - into <strong>cryptographic proofs</strong>. These proofs can then be used anywhere in Web3 to verify that a user is real, reputable, and active, while keeping privacy intact.</p><p>The <a href="http://portal.zkpass.org/"><strong>Verifiable Reputation Score (VRS)</strong></a><strong> </strong>is the <a href="https://github.com/zkPassOfficial/Verifiable-Reputation-Score/blob/main/VRS_Framework_Specification.md">open framework</a> built on top of this foundation. It standardizes how verifiable data is aggregated, normalized, and weighted to quantify trust in a privacy-preserving way. VRS turns individual proofs into a coherent measure of authenticity and contribution across the internet.</p><h3>1. From Proof to Reputation</h3><p>A proof is binary: it tells us that a specific statement is true or false.<br>Reputation is continuous: it captures how credible, consistent, and valuable a participant is across many dimensions.</p><p>The challenge is bridging these two. VRS represents the first open attempt to translate <strong>zero-knowledge proofs of online behavior</strong> into a quantitative reputation metric.</p><p>Instead of relying on self-claimed credentials or centralized scoring services, VRS computes reputation directly from verified user actions across multiple schemas - each representing a data source such as Duolingo, GitHub, Reddit, or Binance. Each proof is verifiable, private, and composable, meaning any dApp can confirm its validity without seeing the raw data.</p><p>This shifts reputation from an <em>opinionated trust system</em> to a <em>mathematical framework</em>. Every score is derived from verifiable evidence, and every weight is transparently defined. In essence, VRS turns the scattered, untrusted signals of the internet into a cryptographic truth layer.</p><h3>2. The Architecture of the VRS Framework</h3><p>The VRS Framework is designed as an <strong>open, modular pipeline</strong>. It transforms proofs into reputation scores through a sequence of standardized operations, ensuring transparency and reproducibility. The architecture consists of five key layers:</p><ol><li><strong>Data Schema Layer: </strong>defines how a Web2 data source is modeled, including attributes, signal types, and expected proof structure.</li><li><strong>Proof Generation Layer:</strong> powered by zkPass’s zkTLS SDK, allows users to locally generate zero-knowledge proofs for their accounts without any intermediaries.</li><li><strong>Signal Extraction Layer: </strong>interprets verified proofs to extract standardized metrics, such as activity frequency, longevity, or verification status.</li><li><strong>Scoring Engine: </strong>computes the Verifiable Reputation Score by combining weighted signals across multiple schemas, applying penalties for incomplete coverage, and normalizing results.</li><li><strong>Verification and Access Layer:</strong> exposes the scores through open APIs or smart contract attestations, enabling any dApp or protocol to consume them.</li></ol><p>Each component is open-source and designed for interoperability. Developers can contribute new schemas, tune scoring weights, or build applications that consume VRS as a trust oracle.</p><h3>3. Data Schemas and Signal Design</h3><p>The foundation of VRS lies in its <strong>Schema Market</strong>, an open registry of verifiable data sources. Each schema defines how a particular Web2 interaction can be verified cryptographically.</p><p>For example:</p><ul><li><strong>Duolingo Schema:</strong> Verifies learning streaks, lesson completions, or course progress.</li><li><strong>LinkedIn Schema:</strong> Confirms verified employment or education records.</li><li><strong>CEX Schema:</strong> Verifies proof of KYC completion or asset holdings without revealing balances.</li><li><strong>GitHub Schema:</strong> Attests to active contribution history and commit frequency.</li><li><strong>Reddit or X Schema:</strong> Confirms long-term account existence and meaningful social engagement.</li></ul><p>Each schema contributes one or more <strong>signals</strong>, such as “Activity Frequency,” “Account Longevity,” or “Verification Depth.” These are then weighted based on reliability, coverage, and resistance to manipulation.</p><p>This modular design allows new data sources to be added seamlessly. If a new schema is published on the zkPass Schema Market, it can immediately be incorporated into the VRS calculation framework. Over 100 schemas have already been integrated, spanning education, identity, social, and finance domains.</p><h3>4. Scoring Methodology</h3><p>The core of VRS lies in its scoring algorithm. It transforms multiple verified signals into a single, normalized score ranging from 0 to 100.</p><p>The process follows a multi-step pipeline:</p><h4>Step 1: Signal Normalization</h4><p>Each raw signal is normalized to a 0–1 range using domain-specific thresholds. For example, a Duolingo streak of 60 days might map to 0.8, while a GitHub commit frequency in the top 10% maps to 1.0. Normalization ensures comparability across heterogeneous data sources.</p><h4>Step 2: Weight Assignment</h4><p>Each signal is assigned a <strong>signal weight (wᵢ)</strong> that reflects its reliability, importance, and resistance to gaming. Identity-based proofs (e.g., verified CEX KYC) typically carry higher weights than activity proofs (e.g., Reddit engagement). All weights are open and defined in the framework’s configuration file.</p><h4>Step 3: Coverage Penalty</h4><p>To prevent users from earning high scores through only one data source, VRS introduces a <strong>coverage-penalized mechanism</strong>. This penalty decreases the composite score when a user has few active schemas. The intuition is that diverse verifiable behaviors indicate stronger reputation credibility.<br>Mathematically, if a user has verified <em>n</em> schemas out of <em>N</em>, the coverage coefficient adjusts total weight proportionally, ensuring fairness and discouraging narrow gaming.</p><h4>Step 4: Aggregation</h4><p>The normalized signals and weights are aggregated using a weighted mean function adjusted by the coverage coefficient:</p><pre>VRS = (Σ wᵢ × sᵢ) × C(n)</pre><p>where sᵢ represents each signal score, and C(n) is the coverage coefficient.</p><h4>Step 5: Calibration</h4><p>Finally, the raw score is calibrated to align distribution across the population. This calibration maintains consistency between cohorts (for example, preventing overrepresentation from any single platform). It also allows VRS to evolve dynamically as more schemas and participants join.</p><p>The result is a transparent, reproducible, and privacy-preserving reputation metric that reflects the true diversity and authenticity of user actions.</p><h3>5. Verifiable Integrity: zkTLS at the Core</h3><p>The reliability of any scoring system depends on the authenticity of its inputs. VRS solves this through <strong>zkTLS</strong>, zkPass’s proprietary zero-knowledge transport layer protocol.</p><p>zkTLS allows users to prove that a specific HTTPS session between their browser and a Web2 platform occurred, without revealing any session content or credentials. It captures and verifies the <strong>TLS handshake</strong> cryptographically, proving the origin and integrity of the data while maintaining complete privacy.</p><p>This means:</p><ul><li>No third-party intermediary ever accesses user credentials.</li><li>No centralized data collection occurs.</li><li>Every proof carries mathematical integrity verified against the website’s genuine TLS certificate.</li></ul><p>By embedding zkTLS into every proof generation process, VRS ensures that the signals it consumes are both authentic and privacy-preserving. There are no forged screenshots, no fake APIs, and no unverifiable claims. Every reputation point is grounded in cryptographic truth.</p><h3>6. Open Framework and Community Extensibility</h3><p>The VRS Framework is designed as a <strong>public infrastructure</strong>, not a closed product. It operates on open standards that any developer, researcher, or protocol can extend.</p><p><strong>a. Schema Contribution:</strong><br>Anyone can design and publish new schemas through the zkPass DevHub, defining the proof logic, expected fields, and normalization parameters. Once approved, the schema becomes available to the community through the Schema Market.</p><p><strong>b. Custom Scoring Extensions:</strong><br>Projects can fork the open scoring framework, adjust signal weights, or plug in domain-specific logic (e.g., for DeFi credit scoring or social engagement analysis). The reference implementation on GitHub serves as a transparent baseline.</p><p><strong>c. API and SDK Access:</strong><br>VRS exposes its computation pipeline through APIs and SDKs, enabling external applications to query verifiable reputation data or build custom leaderboards, eligibility checks, or reward systems.</p><p><strong>d. Governance and Transparency:</strong><br>All weights, formulas, and schema definitions are open-source. The goal is to make VRS a community-governed framework where contributions and audits keep the system fair, neutral, and verifiable.</p><p>This openness turns VRS into a public good — a shared scoring layer that enhances reputation across the entire Verifiable Internet.</p><h3>7. Applications Across the Verifiable Economy</h3><p>The potential applications of VRS are vast, touching nearly every sector of the emerging proof-based economy.</p><p><strong>1. Launchpad Allocation and Fair Distribution</strong><br>In token sales and community launches, reputation should matter more than speed or capital size. VRS enables <strong>reputation-weighted allocations</strong>, where participants with verifiable contributions or authentic activity receive higher priority. The Kaito Capital Launchpad pioneered this by integrating VRS to reward real provers instead of bots or airdrop hunters.</p><p><strong>2. Airdrop Sybil Resistance</strong><br>By verifying diverse schemas, users build multi-dimensional proof profiles that are almost impossible to fake. Airdrops using VRS can reward real, engaged users without invasive KYC or opaque heuristics.</p><p><strong>3. Under-Collateralized Lending</strong><br>Financial protocols can use VRS to assess creditworthiness without requesting private data. A verified identity, consistent activity, and on-chain reliability together form a new model for decentralized credit.</p><p><strong>4. Proof of Learning and Proof of Skill</strong><br>Educational and professional platforms can integrate schemas like Duolingo, Coursera, or LinkedIn to measure verified effort and expertise. VRS makes “learning as an asset” possible.</p><p><strong>5. Reputation-Driven Advertising and Rewards</strong><br>Projects can design incentive programs that target real users with provable credibility, replacing arbitrary impressions with verifiable engagement.</p><p><strong>6. DAO Governance and Voting</strong><br>Reputation-weighted voting models built on VRS ensure that governance power reflects verified contribution, not just token holdings.</p><p>Together, these applications illustrate VRS’s role as the connective tissue of a new verifiable economy - one where privacy and proof coexist seamlessly.</p><h3>8. Privacy, Fairness, and Ethics</h3><p>Reputation systems can easily become surveillance systems if built without restraint. VRS is designed to avoid that outcome entirely.</p><p>The framework never stores or exposes raw data. Every input is a zero-knowledge proof generated locally by the user. Scores are computed from verifiable facts, not from scraped behavior.</p><p>Moreover, users retain full sovereignty over their proofs. They can choose when and where to use their VRS, and no central authority can unilaterally track or modify it.</p><p>Fairness is built into the architecture:</p><ul><li><strong>Open weighting logic</strong> prevents hidden biases.</li><li><strong>Coverage penalty</strong> discourages manipulation through narrow activity.</li><li><strong>Open-source code</strong> allows public audits and community governance.</li></ul><p>In contrast to opaque Web2 reputation systems built on black-box algorithms, VRS is transparent, decentralized, and cryptographically fair. It proves that reputation can be both private and accountable.</p><h3>9. Conclusion</h3><p>The Verifiable Internet cannot rely on belief; it must rely on proof.<br>VRS transforms the concept of online reputation from a social construct into a cryptographic standard. Built by zkPass and powered by zkTLS, it provides the first open, privacy-preserving, and extensible framework for quantifying credibility across the digital world.</p><p>As Web2 and Web3 continue to merge, the ability to prove authenticity without revealing identity will define the next decade of the internet. VRS stands at the center of that transformation.</p><p><strong>All is verifiable.</strong><br>And with VRS, reputation is too.</p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a><em> | </em><a href="http://portal.zkpass.org/"><strong><em>Portal</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=31120a3a9329" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/verifiable-reputation-score-vrs-an-open-framework-by-zkpass-31120a3a9329">Verifiable Reputation Score (VRS) — An Open Framework by zkPass</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Introducing the Verifiable Application Acceleration Program (VAAP)]]></title>
            <link>https://medium.com/zkpass/introducing-the-verifiable-application-acceleration-program-vaap-5868dcf5cee5?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/5868dcf5cee5</guid>
            <category><![CDATA[announcements]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Wed, 08 Oct 2025 14:58:45 GMT</pubDate>
            <atom:updated>2025-10-08T14:58:32.003Z</atom:updated>
            <content:encoded><![CDATA[<p>Empowering the next generation of builders with $5.5M in strategic support</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HoviNpQ9V0zp52TlZpwHcA.jpeg" /></figure><p>As the Verifiable Internet begins to take shape, a new wave of applications is emerging where trust is no longer assumed but proven. To accelerate this transformation, zkPass is launching the <strong>Verifiable Application Acceleration Program (VAAP)</strong>.</p><p>The initiative is supported by $5.5M in strategic funding from a private round completed in early 2025, with participation from a global consortium of investors who share zkPass’s vision for a verifiable and privacy-preserving Internet.</p><p>To date, zkPass has raised over $20M in total funding from leading global investors including YZi Labs (ex-Binance Labs), dao5, Hongshan (ex-Sequoia China), Amber Group, UOB Ventures, and Susquehanna International Group (SIG), as well as other strategic backers worldwide, establishing zkPass as one of the most well-capitalized infrastructures in the zero-knowledge ecosystem.</p><h4>What is a Verifiable Application</h4><p>Not another dApp.</p><p>A <strong>verifiable application</strong> is any system that runs on <strong>proofs instead of promises</strong>. Instead of trusting screenshots, disclosures, or centralized APIs, applications consume cryptographic attestations that guarantee both privacy and integrity.</p><ul><li><strong>zkKYC</strong> lets users prove regulatory compliance without exposing personal records.</li><li><strong>Under-collateralized lending</strong> allows borrowers to demonstrate income or account balances without handing over bank statements.</li><li><strong>Prediction markets</strong> settle outcomes on-chain based on verifiable real-world data rather than centralized feeds.</li><li><strong>Financializing learning data</strong> turns a Duolingo streak, a GitHub commit history, or an online certification into a verifiable digital asset.</li><li><strong>Parametric insurance</strong> enables payouts based on weather or flight data proven directly from Web2 sources.</li><li><strong>Reputation-driven advertising</strong> makes campaigns target audiences based on verifiable reputation, not on leaked or harvested profiles.</li></ul><p>Verifiable applications redefine what data means.</p><p>Data is no longer a static entry locked inside a database or exposed to intermediaries.</p><ul><li>It becomes <strong>portable</strong>, able to move across systems with integrity preserved.</li><li>It becomes <strong>private</strong>, revealing only what is necessary and nothing more.</li><li>It becomes <strong>liquid</strong>, usable as an economic primitive in financial, social, and AI-native ecosystems.</li></ul><p>This is the foundation of the <strong>Verifiable Internet</strong>: a world where applications no longer process raw data, they process <strong>truth</strong>.</p><h4>The VAAP Framework</h4><p>The <strong>Verifiable Application Acceleration Program</strong> provides a full-stack framework designed to help teams build, launch, and scale their verifiable products. Participants will gain access to four key pillars of support:</p><ol><li><strong>Technology Enablement</strong><br>Access zkPass SDKs, DevHub, and Schema Market incentives to integrate verifiable proofs quickly and efficiently.</li><li><strong>Go-to-Market Support</strong><br>Leverage zkPass’s partner network, co-marketing campaigns, and ecosystem distribution channels to accelerate user acquisition and visibility.</li><li><strong>Enterprise and Compliance Guidance</strong><br>Gain direct connections with exchanges, financial institutions, and advisory partners to support regulated or enterprise-grade adoption.</li><li><strong>Funding Alignment</strong><br>Projects demonstrating strong proof-of-business can access co-investment opportunities from zkPass and its strategic backers.</li></ol><h4>Commitments for VAAP Applications</h4><p>Each selected project becomes a contributing participant in the zkPass ecosystem. In joining VAAP, teams are expected to:</p><ul><li><strong>Build for real adoption</strong>: design products with clear business models that prove the value of verifiable data</li><li><strong>Share protocol revenue</strong>: allocate a portion of application income back to the zkPass protocol to sustain long-term growth</li><li><strong>Adopt $ZKP</strong>: integrate the $ZKP token into settlement, access, or staking mechanisms so that application activity directly reinforces token utility</li><li><strong>Follow verifiable-first design</strong>: ensure that proofs, not raw data, form the foundation of trust and user experience</li></ul><h4>Momentum and Vision</h4><p>Over the past two years, zkPass has become the leading verifiable data infrastructure in Web3.</p><ul><li>More than <strong>1 million unique attested users</strong> across ecosystem integrations</li><li>Over <strong>200 million total verifiable computations (TVC)</strong></li><li><strong>80+ partner integrations</strong> spanning DeFi, Social, Gaming, DePIN, and AI</li><li><strong>100+ verified data sources</strong> including Binance, Coinbase, Duolingo, Uber, LinkedIn, and more</li></ul><p>zkPass is standardizing credibility as a network primitive for the Verifiable Internet. The launch of VAAP marks the next phase of this movement, empowering builders to turn verifiable proofs into meaningful economic systems.</p><p>Applications are now open for global partners and developers.</p><p><strong>Apply here:</strong> <a href="https://form.typeform.com/to/z1IE80Nx">https://form.typeform.com/to/z1IE80Nx</a></p><p><strong>Make the world verifiable.</strong></p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a><em> | </em><a href="http://portal.zkpass.org/"><strong><em>Portal</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5868dcf5cee5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/introducing-the-verifiable-application-acceleration-program-vaap-5868dcf5cee5">Introducing the Verifiable Application Acceleration Program (VAAP)</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Onchain Equities Are Not Just Tokens — They’re an Entry Point to Financial Recomposition]]></title>
            <link>https://medium.com/zkpass/onchain-equities-are-not-just-tokens-theyre-an-entry-point-to-financial-recomposition-1bf51b1009b2?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/1bf51b1009b2</guid>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Wed, 02 Jul 2025 14:20:28 GMT</pubDate>
            <atom:updated>2025-07-02T14:20:28.660Z</atom:updated>
            <content:encoded><![CDATA[<h3>Onchain Equities Are Not Just Tokens — They’re an Entry Point to Financial Recomposition</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YsQ4mDUq3HhHV1cCc7Lv-w.png" /></figure><p>In 2025, “tokenized stocks” have returned to the crypto narrative spotlight. But this time, the story runs deeper than the old, tired refrain of “real-world assets onchain.”</p><p>This is no longer about mapping traditional finance onto crypto rails.</p><p>It’s about re-architecting the language of financial systems:</p><ul><li>Not just mirroring assets, but migrating structure.</li><li>Not just issuing tokens, but encoding permission, composition, and self-sovereignty.</li><li>Not replicating the old world, but creating a new grammar for capital.</li></ul><p>And quietly embedded in this process, <strong>zkTLS</strong> may not be the protagonist — but it could be the one component that makes the entire thing verifiable.</p><h4>Tokenized stocks aren’t RWA 2.0 — they might be DeFi’s native asset class</h4><p>RWA has largely been built on the premise of <em>certainty</em>: bonds, treasuries, invoices, cash equivalents — assets that lend themselves well to wrappers, vaults, and yield protocols.</p><p>Stocks are a different beast.</p><p>They are <strong>volatile, narrative-driven, and structurally dynamic</strong>. They don’t just <em>carry value</em> — they <em>create interaction</em>. Which means:</p><blockquote><em>Once equities become programmable assets, they don’t merely integrate into DeFi — they redefine what DeFi </em>is<em>.</em></blockquote><p>Robinhood, Backed, Swarm, and Superstate aren’t just issuing mirrored assets. They’re building systems in which $TSLA becomes:</p><ul><li>A composable module in a leverage vault,</li><li>A yield-bearing LP with USDC,</li><li>A collateral asset for onchain credit,</li><li>A signal for identity or governance participation.</li></ul><p>This is not tokenization. It’s <strong>financial recomposition</strong> — rewriting what it means for an asset to exist and act.</p><h4>The hard part was never issuance — it’s execution</h4><p>All of this rests on a fragile assumption: that you can <strong>trust the data</strong>.</p><p>How do you know a tokenized AAPL position corresponds to a real Robinhood account?<br> How can an onchain contract know that a dividend has been announced?<br> How can a vault verify a stock split has occurred — or trigger a rebalance based on it?</p><p>Worse yet: how can you confirm someone is a qualified investor — <strong>without exposing any personal information</strong>?</p><p>Most RWA systems today rely on a predictable formula: custodians + APIs + trusted oracles. But the moment you try to <strong>automate</strong>, <strong>compose</strong>, and <strong>scale</strong>, this framework breaks.</p><p>And this isn’t just a philosophical dilemma. It’s a systemic bottleneck.</p><h4>zkTLS is not a headline — it’s a structural hinge</h4><p>zkTLS solves a very specific problem:</p><blockquote><em>It turns HTTPS web data into onchain zero-knowledge proofs.</em></blockquote><p>It doesn’t make APIs redundant. It doesn’t compete with Chainlink. It doesn’t promise to “tokenize everything.”</p><p>But it quietly answers a question nobody else can:</p><blockquote>“Can this smart contract verify that a human truly retrieved this data from sec.gov?”</blockquote><p>In the context of onchain equities, that’s transformative.</p><p>It enables:</p><ul><li><strong>Automated corporate action triggers</strong>: Verifiably fetching and proving dividend, stock split, or merger announcements from official sources.</li><li><strong>zk-based compliance</strong>: Proving user identity or investor qualification without revealing any KYC data.</li><li><strong>Proof of holdings</strong>: Extracting a snapshot of a TradFi account to enable onchain mirroring or token minting.</li><li><strong>Strategy inputs</strong>: Allowing vaults or structured products to ingest live, verified external conditions.</li></ul><p>It doesn’t ask you to get excited about it. But it makes sure that excitement won’t collapse under unverifiable inputs.</p><h4>The bigger picture: DeFi can’t scale without trustless data</h4><p>For five years, DeFi has optimized for code over trust — but that optimization only works in closed systems.<br> As RWA expands, the data layer becomes the new point of failure.</p><p>zkTLS doesn’t solve everything. It’s not a general oracle. But it introduces a missing pillar: <strong>web-native data with cryptographic verifiability</strong>.</p><p>That alone is enough to unlock new forms of automation, compliance, and execution logic — especially when the data is sensitive, high-stakes, or legally consequential.</p><h4>Final thoughts</h4><p>The tokenization of equities isn’t about minting stock-pegged tokens. It’s a wedge into rethinking financial agency, programmability, and system design.</p><p>And zkTLS isn’t the story’s main character. It’s a background detail — a component you might never notice.</p><p>But if you’re serious about building the next generation of onchain finance — especially systems that ingest real-world signals and trigger real-world outcomes — you’ll need a way to trust what you read from the web.</p><blockquote><em>Not because someone told you so. But because you can prove it.</em></blockquote><p>If you’re building in this space — structured products, composable compliance, agentic finance, or data-integrated vaults — now is the time to rethink the trust layer.</p><p><strong>zkTLS isn’t the narrative. But it may be the missing verification.</strong></p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a><em> | </em><a href="http://portal.zkpass.org/"><strong><em>Portal</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1bf51b1009b2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/onchain-equities-are-not-just-tokens-theyre-an-entry-point-to-financial-recomposition-1bf51b1009b2">Onchain Equities Are Not Just Tokens — They’re an Entry Point to Financial Recomposition</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[VOLE-In-the-Head (VOLEitH): the superpower to scale zkTLS]]></title>
            <link>https://medium.com/zkpass/vole-in-the-head-voleith-the-superpower-to-scale-zktls-0e84a5b13e71?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/0e84a5b13e71</guid>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Benjamin]]></dc:creator>
            <pubDate>Tue, 10 Jun 2025 06:03:10 GMT</pubDate>
            <atom:updated>2025-06-10T06:05:16.158Z</atom:updated>
            <content:encoded><![CDATA[<p>VOLE-In-The-Head(VOLEitH): build a Better, Faster, and More Secure zkTLS</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GZBzdCB95PcTdP_m-bAxTQ.png" /></figure><h3>Introduction: zkPass and the Shift to VOLEitH</h3><p>Imagine a world where you can prove something about yourself — like your age or credit score — without ever showing the actual details. That’s what <strong>zkPass</strong> does. It’s a privacy-first protocol built on zero-knowledge proofs, or ZKPs, letting you keep your data secure while still proving what matters. Originally, zkPass mixed VOLE-based ZK with zk-SNARKs, but now it’s made a big leap: it’s fully transitioned to the VOLE-In-The-Head(VOLEitH) protocol. Why? Because VOLEitH boosts security, speeds things up, and fits perfectly with the decentralized spirit of Web3. Let’s break it down.</p><h3>Why VOLEitH? The Big Wins</h3><p>So, what makes VOLEitH so special for zkPass? Here are the key advantages:</p><p>1. <strong>No Trusted Setup</strong></p><p>- Some systems, like zk-SNARKs, need a trusted setup — a process that could hide security risks if it’s not done right. VOLEitH skips that entirely. No trust needed, just pure security and a simpler setup.</p><p>2. <strong>Anyone Can Verify</strong></p><p>- With VOLEitH, proofs are public. Anyone can check them using openly available info — no special verifiers required. That’s a game-changer for transparency in decentralized systems, unlike older protocols that kept things more closed-off.</p><p>3. <strong>Fast and Light</strong></p><p>- VOLEitH is built to be efficient. It generates and verifies proofs quickly, without hogging memory or computing power. That’s critical for real-world apps where speed matters.</p><p>4. <strong>Privacy Stays with You</strong></p><p>- Here’s the best part: you generate proofs right on your own device. Your sensitive data never leaves your hands — no middlemen, no escrow, just you in control. That’s privacy done right.</p><h3>VOLEitH vs. the Competition</h3><p>How does VOLEitH stack up against other zero-knowledge protocols? Let’s compare:</p><p>1. <strong>VOLEitH vs. zk-SNARKs</strong></p><p>- zk-SNARKs need that risky trusted setup — VOLEitH doesn’t. Plus, VOLEitH’s public verifiability beats zk-SNARKs’ more restricted approach. It’s safer, simpler, and more open.</p><p>2. <strong>VOLEitH vs. zk-STARKs</strong></p><p>- zk-STARKs also skip the trusted setup, but VOLEitH shines for client-side use. It leans on symmetric cryptography and generates proofs in linear time, making it lighter and faster for your device.</p><h3>Why VOLEitH is a Perfect Match for zkTLS</h3><p>Now, let’s talk about <strong>zkTLS</strong> — a system that securely bridges Web2 data to Web3. VOLEitH is tailor-made for it. Here’s why:</p><p>1. <strong>Public Trust</strong></p><p>- zkTLS needs proofs that the whole Web3 world can trust. VOLEitH’s public verifiability delivers that without relying on any central authority.</p><p>2. <strong>No Setup Hassles</strong></p><p>- Without a trusted setup, VOLEitH makes zkTLS more secure and easier to roll out across decentralized networks.</p><p>3. <strong>Efficient for the Job</strong></p><p>- zkTLS often deals with straightforward, linear evaluations. VOLEitH handles these better than zk-SNARKs, cutting down on computing power and keeping things snappy.</p><p>4. <strong>Client-Friendly</strong></p><p>- zk-SNARKs can be sluggish and heavy on devices, even for basic zkTLS tasks. VOLEitH? It’s simpler, smaller, and faster — perfect for generating proofs locally without breaking a sweat.</p><h3>The Future: Making VOLEitH Even Better</h3><p>zkPass isn’t stopping here. We are already working on upgrades:</p><p>1. <strong>Speed Boost</strong></p><p>- A technique called “One-Tree-To-Rule-Them-All” will slash proof generation times and cut computational costs even further.</p><p>2. <strong>Smaller Proofs</strong></p><p>- A recursive verifier will shrink proof sizes, making them ideal for on-chain verification — think blockchain-ready efficiency.</p><h3>Summary</h3><p>By switching to VOLEitH, zkPass has leveled up. It’s more secure without trusted setups, more transparent with public proofs, and more efficient for real-world use. For zkTLS and beyond, VOLEitH is the perfect fit — keeping your data private, your proofs fast, and Web3 truly decentralized.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0e84a5b13e71" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/vole-in-the-head-voleith-the-superpower-to-scale-zktls-0e84a5b13e71">VOLE-In-the-Head (VOLEitH): the superpower to scale zkTLS</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[zkTLS — The Cornerstone of Verifiable Internet]]></title>
            <link>https://medium.com/zkpass/zktls-the-cornerstone-of-verifiable-internet-da8609a32754?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/da8609a32754</guid>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Mon, 16 Dec 2024 16:48:00 GMT</pubDate>
            <atom:updated>2024-12-16T16:37:20.972Z</atom:updated>
            <content:encoded><![CDATA[<h3>zkTLS — The Cornerstone of Verifiable Internet</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sdPW5Q7peRX63alXPaZWDA.png" /></figure><p>The Verifiable Network represents a new paradigm for the internet, empowering users to verify the source and authenticity of network data in a decentralized manner. At its core lies the zkTLS infrastructure, a groundbreaking mechanism for cryptographically verifying any data exchanged over HTTPS — unlocking the true potential of a Verifiable Network.</p><h3>1. Fortress</h3><p>Today’s internet resembles a sprawling web of “fortresses” — various online platforms controlled by centralized entities. These fortresses dictate the rules and decide whether information can flow between platforms. To access services, users must often surrender personal data to these platforms, even when it involves sensitive private information.</p><p>However, the fate of this data is largely opaque. Platforms may share user data with regulatory bodies or third parties without clear disclosure or explicit consent. How these third parties use the data is even harder to track. Many have experienced the irritating outcome of such leaks firsthand — unsolicited sales calls are often the direct result of personal data being misused.</p><p>Blockchain technology offers inherent advantages in establishing data ownership, but the volume of on-chain data remains limited compared to traditional internet systems. Decentralized applications (dApps) face significant hurdles in accessing real-world data, as data silos severely constrain their development potential and limit the scope of blockchain innovation.</p><p>Oracles serve as the bridge between the crypto world and the real world, making them a critical enabler for decentralized technologies.</p><h3>2. Oracle</h3><p>In blockchain systems, an Oracle is a service that brings external data into blockchain smart contracts. Since blockchains themselves cannot access off-chain data, oracles act as a bridge between the blockchain and the outside world, enabling smart contracts to execute actions based on external information.</p><p>The primary functions of an Oracle include:</p><ul><li><strong>Data Transmission</strong>: Delivering external data, such as financial market prices, weather information, or sports results, to the blockchain.</li><li><strong>Data Verification</strong>: Ensuring that the transmitted data is accurate and trustworthy.</li><li><strong>Triggering Smart Contracts</strong>: Activating the execution of smart contracts based on changes in external data.</li></ul><h4>2.1 Public Data Oracle</h4><p>Public data refers to information that is freely accessible and available for use.</p><p>The crypto ecosystem has successfully bridged Web2 public data through Oracles such as <strong>Chainlink</strong>, <strong>Pyth Network</strong>, <strong>Supra</strong>, <strong>UMA</strong>, <strong>Tellor</strong>, and <strong>Redstone</strong>. Public data is critical for advancing <strong>DeFi </strong>on blockchain by providing transparency and enabling automation. Without these data sources, the capabilities of decentralized finance would be significantly constrained.</p><p>Because the data sources are public, this category of oracles is referred to as <strong>Public Data Oracles</strong>.</p><p>However, they represent only a fraction of the potential:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lX33XPFJ2GUtNjlG" /></figure><p>In Web3, private data remains largely inaccessible, posing significant barriers to the development and adoption of consumer applications and real-world use cases. We believe the future internet should go beyond DeFi’s dominance in sectors like finance and enable a broader range of impactful applications.</p><h4>2.2 Private Data Oracle</h4><p>Private data refers to protected information that requires authorization to access.</p><p>Think of private data as a personal diary of your life — it is tailored to you and only you. This includes data that identifies who you are or reveals intimate details about your life, such as your social security number, bank account details, health records, or other uniquely personal information. While public data drives informed decisions, transparency, and research, private data — when handled with respect and legal safeguards — can unlock personalized services while protecting individual privacy.</p><p>Over decades of Web2 evolution, private data has become the new <strong>digital oil</strong> of our age.</p><p>The true game-changer lies in unlocking private data — rich, sensitive information. When securely accessed (using cryptographic methods that empower users with control, trust, and privacy), private data has the potential to revolutionize the next wave of consumer applications.</p><p><strong>zkTLS is the key to unlocking private data.</strong></p><h3>3. TLS</h3><p>To understand zkTLS, we first need to explore the TLS protocol — a foundational technology of the internet that most of us interact with daily, often unknowingly.</p><p>Remember the “fortresses” of the internet? The pathways connecting users to these fortresses are primarily built on the HTTPS protocol (HyperText Transfer Protocol Secure). While you may not be familiar with the technical details, you’ve certainly typed something like <a href="https://www.123.com/">https://www.123.com</a> into your browser’s address bar. That’s HTTPS — a protocol that enables your browser to interact with internet platforms.</p><p><strong>What is TLS?</strong></p><p>TLS (Transport Layer Security) is the core component of HTTPS, operating one layer deeper in the OSI protocol model. TLS ensures the confidentiality and integrity of data during transmission through encryption, authentication, and integrity protection, safeguarding against data theft or tampering. It is one of the foundational and mature technologies underpinning today’s internet.</p><p>You don’t need to dive into the technical details to understand its role — it’s akin to deploying sentries along the pathways to the fortresses. When their passphrases (Handshake) match, they ensure the safe passage of messages between you and the fortress.</p><p><strong>The Limitations of Traditional TLS</strong></p><p>TLS is essentially a two-party protocol, designed to establish a secure connection between a client and a server. In traditional TLS, the integrity and privacy of data are guaranteed only between these two parties. However, these fortress-like servers are highly guarded, often unwilling to share user data with third parties, leading to the formation of data silos.</p><p>Even if a data source is willing to make its data assets available to third parties, the standard two-party TLS protocol struggles to provide proof of this data’s integrity and authenticity to others. Simply forwarding an HTTPS response from the server to a third party isn’t convincing, as the client can easily tamper with the data locally. Moreover, forwarding such responses completely exposes them to third parties, compromising user privacy — something no user would want.</p><p><strong>Challenge</strong></p><p>The key challenge, therefore, is finding a way to prove the <strong>integrity and authenticity of data</strong> to any third party while preserving <strong>privacy</strong>. Achieving this requires overcoming both technical and trust-related hurdles, and this is precisely where zkTLS steps in to provide a solution.</p><h4>3.1 RFC9421</h4><p><a href="https://datatracker.ietf.org/doc/html/rfc9421">RFC9421</a> attempted to address this challenge but faced significant hurdles in its implementation. It requires coordination across multiple parties, including certificate issuance by authoritative bodies and synchronized modifications to both client and data source servers. This extensive involvement of various stakeholders makes real-world deployment exceptionally challenging.</p><p><strong>Logical Limitations of RFC 9421</strong></p><p>From a logical standpoint, the design philosophy of RFC 9421 does not align with the practical needs of existing data interaction systems, rendering it somewhat redundant. A data source has no obligation to prove the origin of data to a third party. Its responsibility is solely to the direct requester, ensuring the integrity and authenticity of the data provided.</p><p>If a third party requires verification of data provenance, that is entirely their own need, independent of the data source. When a user retrieves information from a data source and chooses to “forward” it to a third party, this is a personal decision by the user. The data source is not required to support such indirect relationships, let alone bear additional costs to provide proofs for them.</p><p><strong>Lack of Adoption and Bleak Prospects</strong></p><p>To date, no major company has announced support for RFC 9421, nor has any hinted at plans to adopt it in the future. The application prospects for this protocol are extremely slim. In fact, we pessimistically believe that it is unlikely to ever see real-world adoption within our lifetime.</p><h4>3.2 OAuth</h4><p>The next challenger is <strong>OAuth</strong>, a well-established and mature solution. You don’t need to dive into its technical details, as you’ve likely encountered OAuth in action when logging into other websites using your Google account.</p><p><strong>How OAuth Addresses Data Verification Needs</strong></p><p>When third parties need to verify the origin of data, OAuth can largely fulfill the requirements:</p><ul><li><strong>Authorization and Verification</strong>: OAuth allows the data source to authorize third parties to directly access the data, ensuring the credibility of the data’s origin.</li><li><strong>Simplicity and Efficiency</strong>: With a highly mature ecosystem, OAuth is widely supported and offers strong compatibility, making it far less costly to adopt compared to introducing a new protocol.</li><li><strong>Commercialization Potential</strong>: Beyond solving verification issues, OAuth enables data sources to charge for access, creating an opportunity for monetization.</li></ul><p><strong>The Limitations of OAuth</strong></p><p>However, OAuth is not without its flaws. Beyond the need for identity providers to support it, the adaptation and modification required for data sources, and the high service fees associated with it, <strong>the limited scope of supported data sources</strong> stands out as its biggest bottleneck:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dNPsYo1iHzqcNAA-" /></figure><p>According to authoritative data from <strong>Statista</strong> and <strong>WebsiteSetup</strong>, only tens of thousands of applications and websites currently support OAuth. Meanwhile, billions of valuable private data sources either do not support OAuth or have no interest in doing so.</p><p>So, is there a solution that can:</p><ol><li>Be compatible with the vast majority of Web2 data sources,</li><li>Require no adaptation or modification on the part of data sources,</li><li>Ensure the <strong>privacy</strong>, <strong>integrity</strong>, and <strong>authenticity</strong> of the data?</li></ol><p>Such a solution is exactly what zkTLS aims to deliver — an innovative protocol that overcomes these limitations and unlocks the potential of private data without requiring data source participation or compromising privacy.</p><h4>3.3 MPC-TLS</h4><p>Founded in 2013, the non-profit project <a href="https://tlsnotary.org/">TLSNotary</a> introduced a novel approach to solving the problem: <strong>MPC-TLS</strong>. This method uses <strong>secure multi-party computation (MPC)</strong> to add a verifier to the TLS connection. Through cryptographic guarantees provided by MPC, the verifier can validate TLS records without requiring external trust or exposing the complete data to anyone.</p><p>We believe that the greatest contribution of TLSNotary lies in its successful engineering and open-sourcing of the concept of introducing a third-party <strong>notary/verifier</strong> using MPC technology, specifically <strong>Garbled Circuits</strong>. This innovation laid the groundwork for the subsequent development of <strong>zkTLS</strong>, advancing the possibilities for privacy-preserving and verifiable data solutions.</p><h3>4. zkTLS</h3><p><strong>zkTLS</strong> (Zero-Knowledge Transport Layer Security) is a multi-party protocol (typically involving three parties) that creates a gateway between Web2 private data and the Web3 ecosystem. zkTLS is a hybrid protocol that integrates <strong>ZKP (Zero-Knowledge Proofs)</strong> with the <strong>TLS encryption system</strong>, enabling the secure transmission of data between Web2 and Web3 while preserving privacy. It allows users to securely extract data from any website without the risk of leaking unnecessary information.</p><p><strong>Key Components of zkTLS</strong></p><p>As the name suggests, zkTLS integrates and is compatible with two major technologies:</p><ol><li><strong>TLS (Transport Layer Security)</strong>:<br>TLS is an encryption protocol that underpins secure web communications on the internet. It is widely used for securely browsing data and encrypting information transmitted between browsers and web servers.</li><li><strong>ZK (Zero-Knowledge Proofs)</strong>:<br>ZKP allows a prover to demonstrate the validity of information to a verifier without revealing the actual content of the information, ensuring data privacy.</li></ol><p><strong>Building a zkTLS Protocol: Two Essential Steps</strong></p><p><strong>1. Establishing a TLS Session with Notarized Security and Integrity</strong>:<br>This involves creating a multi-party TLS session protocol that meets standard TLS requirements while enabling notarization. Multi-party TLS sessions can be constructed using approaches such as:</p><ul><li>Trusted Execution Environments (TEE)</li><li>Secure Multi-Party Computation (MPC)</li><li>Proxy-based models</li></ul><p><strong>2. Generating Zero-Knowledge Proofs for the TLS Session</strong>:<br>This step ensures the privacy of the data by converting the TLS session into a zero-knowledge proof. This can be achieved using various zero-knowledge algorithms, including <strong>interactive</strong> or <strong>non-interactive zero-knowledge proofs</strong>.</p><p><strong>Classification by Participants and ZK Types</strong></p><p>The distinctions in zkTLS can be roughly categorized based on:</p><ul><li>The parties involved in the TLS session, and</li><li>The types of Zero-Knowledge proofs applied.</li></ul><p>This framework allows for flexibility in implementation while maintaining the core principles of security, privacy, and interoperability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rvRBGHuiURXyJ6nZ" /></figure><h4>4.1 TEE Model</h4><p><strong>4.1.1 Introduction</strong></p><p>Trusted Execution Environment (TEE) is not a new technology; it has existed in traditional computing for decades. A TEE is a specialized, nearly tamper-proof component of modern CPUs designed to securely execute sensitive computations within an <strong>enclave</strong>, ensuring the integrity and confidentiality of the data and code it processes. TEEs can safeguard sensitive operations, such as user authentication, data decryption, or cryptographic signing, even when the rest of the system is compromised. This isolation enables secure computation without exposing sensitive information to service providers or external parties, making it ideal for privacy-critical processes.</p><p>In a TEE-based model, the prover passes the raw data values obtained from a server into the TEE. The TEE validates the message content within the enclave, performs authentication, and encrypts the message (creating a network proof), which is then transmitted directly to the third-party protocol requesting verification. This model ensures the authenticity of a website’s response.</p><p>In a TEE-based zkTLS setup, the TEE securely runs the protocol, providing proof of the TLS session execution and its content. Here, the verifier is the TEE itself, meaning the system’s trust model hinges on the reputation of the TEE manufacturer or service provider and the TEE’s resistance to physical or side-channel attacks. For users willing to accept the inherent trust and security risks associated with TEEs, this approach to zkTLS is highly efficient, with minimal computational and network overhead.</p><p><strong>4.1.2 Trust Assumptions</strong></p><ul><li>Trust in the hardware provider.</li><li>Acceptance of the risks posed by potential side-channel attacks.</li></ul><p><strong>4.1.3 Advantages</strong></p><ul><li>No need for third-party notaries or proxies to verify the data.</li><li>Minimal network overhead.</li></ul><p><strong>4.1.4 Disadvantages</strong></p><ul><li>Heavy reliance on hardware, reducing flexibility and exposing the system to hardware-specific vulnerabilities.</li><li>Limited availability across different devices, which may hinder adoption.</li></ul><p><strong>4.1.5 Key Adopters</strong></p><p><a href="https://tech.cornell.edu/built/town-crier/"><strong>Town Crier</strong></a>: Created by <a href="https://www.initc3.org/">IC3</a> in 2017 and acquired by Chainlink in 2018, Town Crier was the first Oracle supported by a TEE. Programs like Town Crier run within an enclave, which ensures:</p><ul><li><strong>Integrity</strong>: Attackers cannot tamper with the code inside the enclave, guaranteeing the Oracle’s honesty.</li><li><strong>Confidentiality</strong>: Attackers cannot access the content within the enclave.</li></ul><p><a href="https://www.clique.tech/"><strong>Clique</strong></a>: A protocol called <strong>TEE-TLS</strong> enables clients to securely and verifiably call off-chain TLS sessions directly from smart contracts, end-to-end. For sensitive data such as keys and OAuth tokens, encryption is performed within the TEE enclave to ensure complete privacy throughout the process.</p><p><strong>4.1.6 Summary</strong></p><p>TEE is a relatively underdeveloped area in the cryptographic domain, with much of its work still in an experimental stage. Its heavy reliance on hardware poses challenges for large-scale adoption in client applications. This reliance, combined with limited device compatibility, may restrict its broader application.</p><h4>4.2 MPC Model</h4><p><strong>4.2.1 Introduction</strong></p><p>To prove a TLS session, the <strong>Client</strong> connects with a <strong>Notary</strong>, and together they use <strong>Secure Multi-Party Computation (MPC)</strong> to generate a shared public key, which is then sent to the <strong>Server</strong>. This key is used during the TLS handshake to create the <strong>Pre-Master Secret</strong> and subsequently the <strong>Session Keys</strong>.</p><p>The session key is then split into two parts: one held by the Client and the other by the Notary. Essentially, the Notary and Client function as a single, unified “super-client,” which follows the standard TLS protocol for encrypted communication with the Server.</p><p>From the Server’s perspective, this process is entirely transparent, requiring no modifications or adaptations — an important advantage as it avoids imposing additional demands on the Server.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*UGk7_xG2aSax_Chg" /></figure><p>At the end of the session, the <strong>Client</strong> creates an authenticated commitment to the plaintext data. The <strong>Notary</strong> verifies this commitment and then signs it without ever seeing the actual data. Once this step is completed, the Client can generate a <strong>ZK proof</strong> that reveals only the necessary information.</p><p><strong>4.2.2 Trust Assumptions</strong></p><ul><li>Assumes the prover does not collude with the third-party Notary/Verifier.</li></ul><p><strong>4.2.3 Advantages</strong></p><ul><li><strong>No risk of IP blocking by the server</strong>: The server interacts with what it perceives as a single client, eliminating the possibility of blocking specific IP addresses.</li></ul><p><strong>4.2.4 Disadvantages</strong></p><ul><li><strong>Latency Sensitivity</strong>: The execution time of <strong>MPC</strong> is highly sensitive to network latency. Even slight increases in delay can significantly impact the total time required for MPC completion, leading to inconsistencies in proof generation.</li><li><strong>Data Size Constraints</strong>: The computational workload and volume of exchanged data during MPC increase dramatically with the size of the data being proved.</li><li><strong>High Performance Overhead</strong>: This is primarily due to the <strong>truth tables</strong> in the Garbled Circuit. Each application data block must be encrypted using a truth table, greatly increasing computation and bandwidth requirements.</li><li><strong>Fixed Costs</strong>: For a single TLS session, there is a fixed cost of <strong>20MB</strong>.</li><li><strong>Additional Costs</strong>: For every <strong>1KB of outgoing data</strong>, an additional <strong>8MB</strong> is required, and for every <strong>1KB of incoming data</strong>, an additional <strong>32KB</strong> is needed.</li><li><strong>Example</strong>: For a typical <strong>1KB request</strong> and a <strong>100KB response</strong>, the prover would need to upload approximately <strong>32MB</strong>, making this approach impractical for regular usage.</li></ul><p><strong>4.2.5 Key Adopters</strong></p><p><strong>TLSNotary</strong> (introduced earlier) and <strong>DECO</strong> represent significant milestones in advancing TLS-based verification.</p><p>Created in 2020 by students and faculty from <strong>IC3</strong>, <a href="https://www.deco.works/">DECO</a> builds upon the foundation laid by TLSNotary. It was acquired by <strong>Chainlink</strong> in the same year.</p><p>DECO enhances TLSNotary’s approach by integrating <strong>Zero-Knowledge Proofs (ZKPs)</strong>, specifically <strong>zk-SNARKs</strong> (as described in its <a href="https://arxiv.org/pdf/1909.00938.pdf">research paper</a>). This addition enables users to prove that a specific piece of data accessed via TLS originates from a designated website.</p><p>Uniquely, DECO also allows users to make selective claims about the data using ZKPs — proving statements about the data without revealing the data itself. This ensures the confidentiality of the underlying information while providing cryptographic guarantees of its authenticity.</p><p><strong>4.2.6 Summary</strong></p><p>The impractical resource requirements of this model severely limit its application in real-world scenarios. However, it proves useful in cases where servers block multiple requests from the same account originating from different IP addresses. In such situations, this model can effectively address the challenge.</p><h4>4.3 Proxy Model</h4><p><strong>4.3.1 Introduction</strong></p><p>The <strong>Proxy Model</strong> in zkTLS leverages the proxy functionality of browsers. Instead of sending requests directly to a website, the client sends them through an HTTPS proxy when opening the site.</p><p>The <strong>Proxy</strong> acts as an intermediary between the client and the server, verifying the encrypted data exchanged during the communication. It does not function as a man-in-the-middle (MitM) because it only has access to encrypted data.</p><p>Once the session is complete, the client shares a portion of the session key that decrypts only the <strong>non-sensitive parts</strong> of the request. The client then generates a <strong>ZKP</strong> for the response data, allowing the proxy and other parties to verify the contents of the TLS session without exposing any sensitive information.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ibLXHd4sz3rxj_bX" /></figure><p>The <strong>Proxy</strong> provides proof of encrypted requests and responses, as well as information indicating whether each piece of data originates from the browser or the website. This effectively proves whether the encrypted data was sent by the browser or the server. Subsequently, the browser generates a <strong>ZK proof</strong> for the decrypted response.</p><p><strong>4.3.2 Trust Assumptions</strong></p><ul><li>Assumes the prover does not collude with the third-party proxy.</li></ul><p><strong>4.3.3 Advantages</strong></p><ul><li><strong>Lower latency</strong> compared to the MPC Model, making it better suited for large-scale, real-time client applications.</li></ul><p><strong>4.3.4 Disadvantages</strong></p><ol><li><strong>Potential bottlenecks</strong>: Overloading the proxy server can cause performance issues.</li><li><strong>Risk of Proxy being banned</strong>: Servers may block requests routed through the proxy.</li><li><strong>High-performance ZK challenges</strong>:</li></ol><ul><li>Implementing high-performance ZK circuits on the client side, especially with tools like Circom, is challenging when dealing with large inputs.</li><li>Website responses often exceed <strong>2KB</strong>, which conservatively requires <strong>8 AES decryptions</strong>. Even if only <strong>1 AES decryption</strong> is required, tests with Circom show an approximate proving time of <strong>5 seconds</strong>, which negatively impacts the user experience on the client side.</li></ul><p><strong>4.3.5 Key Adopter</strong></p><p><strong>Reclaim</strong> Protocol, led by the original Questbook team in 2023, is a proponent of this model. The protocol uses an <strong>HTTP Proxy</strong> to witness the handshake and data transfer. Specifically, it validates:</p><ol><li>The handshake was completed using the correct domain name.</li><li>The correct webpage was accessed in the request.</li><li>The encrypted response was received.</li></ol><p>The encrypted response is then processed by zk-SNARK circuits that run entirely on the client side. AES/ChaCha20 decryption is performed within the ZK circuit, and searches are conducted on the decrypted HTML page.</p><p><strong>4.3.6 Summary</strong></p><p>Most servers do not block access from different IPs, giving this approach relatively broad application potential. For better compatibility with client applications, a low-overhead proxy model requires <strong>memory-efficient ZK algorithms</strong>, given the strict memory limitations of browsers.</p><h4>4.4 Hybrid Mode</h4><p><strong>4.4.1 Introduction</strong></p><p>The vast diversity of clients and data sources means no single model can achieve the broad compatibility required for large-scale commercial adoption. This challenge gave rise to the <strong>Hybrid Mode</strong>, which can be thought of as a combination of a <strong>“high-performance” MPC Model</strong> and a <strong>“memory-efficient” Proxy Model</strong>.</p><p>Key features of the Hybrid Mode include:</p><ol><li><strong>Dynamic Mode Switching</strong>:<br>It adapts to varying characteristics of data sources (e.g., network environments and server policies) by dynamically switching between models to ensure maximum compatibility and an optimal client experience.</li><li><strong>ZK Engineering Balance</strong>:</li></ol><ul><li>Combines <strong>memory efficiency</strong> from interactive Zero-Knowledge Proofs (to handle large requests on browsers with strict memory limits).</li><li>Incorporates the <strong>public verifiability</strong> of non-interactive Zero-Knowledge Proofs (allowing for on-chain verification without requiring real-time verifier availability).</li></ul><p><strong>4.4.2 Trust Assumptions</strong></p><ul><li>Inherits the security assumptions of the existing models (MPC and Proxy).</li></ul><p><strong>4.4.3 Advantages</strong></p><ul><li><strong>Flexible and Adaptive</strong>: Dynamically switches between models, enabling compatibility with a wide range of data sources.</li><li><strong>Enhanced Client Experience</strong>: Ensures the best possible user experience by tailoring the approach to different scenarios.</li></ul><p><strong>4.4.4 Disadvantages</strong></p><ul><li><strong>High Engineering Complexity</strong>: The implementation is technically demanding and requires an extended development timeline.</li></ul><p><strong>4.4.5 Key Adopter</strong></p><p><a href="https://zkpass.org/"><strong>zkPass</strong></a> has pioneered innovations in both the <strong>MPC Model</strong> and the <strong>Proxy Model</strong>:</p><p><strong>2022 — Optimizing the MPC Model</strong>:</p><ol><li>Achieved over <strong>3x efficiency improvements</strong> in MPC algorithms by optimizing communication time, Garbler/Evaluator processes, OT hash functions, and memory copying operations.</li><li>Introduced a new <strong>AES128 proof method</strong>, reducing the block count by <strong>300x</strong> and improving Garbler/Evaluator execution time by <strong>10x</strong>.</li><li><strong>Key improvements:</strong></li></ol><ul><li><a href="https://eprint.iacr.org/2019/448"><strong>Silent OT</strong></a>: Significantly reduced offline network communication during OT operations.</li><li><a href="https://eprint.iacr.org/2020/973"><strong>Stacked GC</strong></a>: Minimized the size of Garbled Circuits, reducing both online communication and execution time.</li></ul><p>These optimizations dramatically shortened the runtime of the entire MPC process.</p><p><strong>2023 — Advancing the Hybrid Model</strong>:</p><ul><li>Introduced <strong>VOLEitH</strong> technology, revolutionizing ZK algorithms.</li><li>Transformed the VOLE-ZK protocol into a <strong>non-interactive format</strong>, enabling fast and efficient proof generation directly in browsers and devices.</li><li>Solved key limitations of traditional SNARKs (high computational and memory demands) and VOLE-ZK (lack of public verifiability).</li></ul><p><strong>4.4.6 Summary</strong></p><p>The Hybrid Mode seamlessly switches between the <strong>Proxy Model</strong> and <strong>MPC Model</strong>, allowing the protocol to adapt to various network environments and server policies without sacrificing efficiency or security. The Proxy Model delivers high performance while minimizing cryptographic overhead, while the MPC Model takes over in scenarios where servers block requests from multiple IP addresses, ensuring privacy and security. This makes it the best practice for zkTLS applications, balancing communication efficiency with client-side computational resources.</p><h3>5. Use Cases</h3><p>In the digital age, <strong>verifiable data</strong> is pervasive, spanning Web3 sectors like DeFi, DAOs, and NFTs, to Web2 realms such as banking, e-commerce, education, healthcare, IoT devices, and social networking. zkTLS is paving the way for a more secure, private, and trustworthy digital ecosystem. In this framework, <strong>privacy and security</strong> are no longer luxuries but fundamental rights for all.</p><h4>5.1 Overall Goal</h4><p>Verifiable data can be categorized into four key objectives:</p><ol><li><strong>Credentials</strong>: Demonstrating private data to any third party.</li><li><strong>Data Integrity</strong>: Enhancing data accuracy and reliability.</li><li><strong>Data Portability</strong>: Seamlessly transferring private data between platforms.</li><li><strong>Data Assets</strong>: Monetizing data as an asset.</li></ol><p><strong>5.1.1 Credentials</strong></p><ul><li><strong>Expanding Airdrop Criteria</strong>: zkTLS extends airdrop eligibility beyond on-chain activity, incorporating user behaviors across traditional internet platforms, overcoming the limitations of conventional airdrops.</li><li><strong>Resisting Sybil Attacks</strong>: Effectively counters bots and Sybil attacks, allowing projects to reward loyal community users instead of opportunistic profiteers.</li><li><strong>On-Chain Credit Scoring</strong>: Addresses the lack of comprehensive credit scores in on-chain lending, reducing collateral requirements and enabling risk-based interest rate models.</li><li><strong>Dynamic Pricing Models</strong>: E-commerce platforms can leverage verifiable purchase histories to offer tailored discounts, attracting high-value customers and fostering healthy competition.</li><li><strong>Automated Verification</strong>: Replaces traditional manual methods like screenshot validation, boosting productivity in scenarios such as verifying applicant details in hiring processes.</li></ul><p><strong>5.1.2 Data Integrity</strong></p><ul><li><strong>Data Quality Control</strong>: In the AI era, high-quality data is essential. zkTLS simplifies operations while ensuring dataset integrity for training models.</li><li><strong>Proof of Humanity</strong>: Distinguishes humans from bots using verifiable data, enhancing platform security.</li><li><strong>Autonomy Proof</strong>: Verifiable data can confirm whether a viewpoint originates from a specific AI agent.</li><li><strong>Filtering Fake Reviews</strong>: zkTLS verifies whether reviewers have genuinely used a product, combating malicious reviews and strengthening platform credibility.</li><li><strong>DePIN Revenue Verification</strong>: Improves transparency and trust in decentralized energy transactions or carbon credit issuance without relying on centralized providers.</li></ul><p><strong>5.1.3 Data Portability</strong></p><ul><li><strong>Ride-Sharing Data Migration</strong>: zkTLS allows seamless transfer of ride history from Platform A to Platform B, enabling new users to gain credibility and discounts immediately.</li><li><strong>Social Graph Import</strong>: Social media platforms can use zkTLS to import users’ social relationships from other networks.</li><li><strong>Reputation Portability</strong>: New platforms can integrate users’ accumulated reputations and followers from other ecosystems, accelerating adoption.</li></ul><p><strong>5.1.4 Data Assets</strong></p><ul><li><strong>Private Data Pools</strong>: As public datasets deplete, AI developers require diverse, personalized data. zkTLS allows private datasets to be verified for authenticity without compromising privacy.</li><li><strong>Data Monetization</strong>: zkTLS enables secure data sharing for personalized offers and discounts, creating win-win scenarios for users and brands.</li></ul><h4>5.2 Application Scenarios</h4><p>Verifiable data enables use cases across various domains, including:</p><ol><li><strong>zkKYC</strong>: Identity verification without revealing sensitive personal details.</li><li><strong>Finance (DeFi &amp; CeFi)</strong>: Integration of Web2 financial information into Web3 ecosystems.</li><li><strong>Governance and Voting</strong>: Privacy-preserving governance and voting solutions.</li><li><strong>Gaming</strong>: Empowering players and developers with multidimensional data.</li><li><strong>Medical and Healthcare</strong>: Secure and private interactions between patients and providers.</li><li><strong>Social Networking</strong>: Enhanced data management and compliance.</li><li><strong>Education and Research</strong>: Secure and private credential verification for students and institutions.</li><li><strong>Experience Checks</strong>: Verifiable background checks for employment.</li></ol><p><strong>5.2.1 zkKYC</strong></p><ul><li><strong>Tenant Verification</strong>: Property managers can verify tenant identities and credit without accessing sensitive information.</li><li><strong>Gaming Age Verification</strong>: Ensures compliance with age restrictions without exposing personal data.</li><li><strong>Social Media Authenticity</strong>: Reduces fake accounts while maintaining user privacy.</li><li><strong>Gig Economy Credentialing</strong>: Verifies freelancers’ identities and qualifications efficiently.</li></ul><p><strong>5.2.2 Finance (DeFi &amp; CeFi)</strong></p><ul><li><strong>Secure Authentication</strong>: zkTLS provides privacy-preserving authentication for DeFi platforms, reducing fraud and boosting trust.</li><li><strong>Credit Verification</strong>: Lending platforms can verify borrowers’ creditworthiness without exposing sensitive data, enabling lower rates for high-credit users.</li><li><strong>Claims Processing</strong>: Insurance companies can verify claims securely and privately.</li><li><strong>Transparent Donations</strong>: Charities can validate donor identities and trace donations without compromising privacy.</li><li><strong>Crowdfunding</strong>: Verifies project creators and supporters securely, fostering trust.</li></ul><p><strong>5.2.3 Governance and Voting</strong></p><ul><li><strong>Privacy-Preserving Voting</strong>: Voters can prove eligibility without revealing their identity.</li><li><strong>Secure Public Service Access</strong>: Citizens can verify residency or citizenship for services without exposing personal details.</li><li><strong>Streamlined Immigration</strong>: Simplifies visa processes with privacy-protected identity verification.</li><li><strong>Shareholder Voting</strong>: Enables secure and anonymous shareholder decision-making.</li></ul><p><strong>5.2.4 Gaming</strong></p><ul><li><strong>Selective Data Sharing</strong>: Players can showcase achievements while protecting private details.</li><li><strong>Universal Gaming Pass</strong>: zkTLS enables seamless login across multiple games, maintaining user control over identity data.</li><li><strong>Achievement Portability</strong>: Allows players to display achievements across ecosystems, fostering a unified gaming reputation.</li></ul><p><strong>5.2.5 Medical and Healthcare</strong></p><ul><li><strong>Secure Health Record Access</strong>: Enables privacy-preserving interaction between patients and providers.</li><li><strong>Telemedicine</strong>: Verifies identities for virtual consultations while safeguarding data.</li><li><strong>Clinical Trials</strong>: Protects participant data while ensuring research transparency.</li><li><strong>Global Care Coordination</strong>: Facilitates secure health data sharing across borders.</li></ul><p><strong>5.2.6 Social Networking</strong></p><ul><li><strong>Decentralized Apps</strong>: zkTLS enhances user trust by enabling selective disclosure of profile details.</li><li><strong>Private Messaging</strong>: Offers fine-grained privacy controls, such as hiding online status or read receipts.</li></ul><p><strong>5.2.7 Education and Research</strong></p><ul><li><strong>Credential Verification</strong>: Simplifies and secures validation of academic achievements.</li><li><strong>Academic Integrity</strong>: Prevents cheating during online exams and fosters secure research collaboration.</li><li><strong>Blockchain-Based Credentials</strong>: Ensures compliance with data protection regulations for cross-border education and research.</li></ul><p><strong>5.2.8 Experience Checks</strong></p><ul><li><strong>Employment Verification</strong>: Employers can confirm candidates’ credentials without exposing sensitive information.</li><li><strong>Transparent Recruitment</strong>: Enables mutual transparency while protecting privacy on job platforms.</li></ul><p><strong>5.2.9 AI Data Labeling</strong></p><ul><li><strong>Secure Data Annotation</strong>: zkTLS ensures labeled datasets are authentic and private, protecting contributors’ sensitive information.</li><li><strong>Collaborative Curation</strong>: Enables pooling of labeled datasets without direct data sharing, fostering secure collaboration.</li><li><strong>Fair Rewards</strong>: Verifies data quality to fairly compensate contributors while maintaining privacy.</li><li><strong>High-Quality Models</strong>: Ensures the integrity of training data, reducing biases and enhancing AI performance.</li></ul><h3>6. Coda</h3><p>In the Web2 era, most data has already been aggregated and now only needs redistribution. zkTLS facilitates this through <strong>trustless</strong> and <strong>selective distribution</strong>.</p><p>As the core infrastructure of the <strong>Verifiable Network</strong>, zkTLS bridges the gap between Web2 and Web3. It sows the seeds of prosperity for on-chain data and opens the door to unprecedented applications. zkTLS is reshaping industries and redefining trust in the digital age.</p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a><em> | </em><a href="http://portal.zkpass.org/"><strong><em>Portal</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=da8609a32754" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/zktls-the-cornerstone-of-verifiable-internet-da8609a32754">zkTLS — The Cornerstone of Verifiable Internet</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Monthly Technical Updates [Nov 2024]]]></title>
            <link>https://medium.com/zkpass/monthly-technical-updates-nov-2024-33d0674008d7?source=rss----1080d7b7c688---4</link>
            <guid isPermaLink="false">https://medium.com/p/33d0674008d7</guid>
            <category><![CDATA[updates]]></category>
            <dc:creator><![CDATA[zkPass]]></dc:creator>
            <pubDate>Fri, 06 Dec 2024 16:06:41 GMT</pubDate>
            <atom:updated>2024-12-06T16:06:41.833Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ykv-vRF8toErUMBwZDI6Pg.png" /></figure><h4>Overview &amp; Conclusion</h4><p>The recent updates focus on improving efficiency, compatibility, and security across systems, with enhanced protocol performance, streamlined user interfaces, and expanded platform integration. These changes aim to address practical challenges, ensure reliability, and support seamless interactions for diverse use cases, paving the way for broader utility and adoption.</p><h4><strong>Protocol Update</strong></h4><ul><li>TLS module refactored to resolve record concatenation and bypass WAFs using updated fingerprints and headers from Chrome and Firefox.</li><li>New forward proxy with WebSocket and WebRTC support launched, enhancing communication.</li><li>ECVRF implemented with secp256k1 and SHA256 for verifiable random node selection.</li><li>Edge tunnel task disperser with WebSocket and WebRTC implemented, using durable objects for better performance.</li><li>V2 API launched with new serialization to reduce bandwidth usage.</li></ul><h4><strong>Transgate Extension</strong></h4><ul><li>UI redesigned for better navigation and data display.</li><li>Protobuf replaced with more efficient serialization, reducing size and improving bandwidth.</li><li>Exploring P2P tech to improve node communication speed and stability.</li><li>Refactored end-of-response check for chunked types for better efficiency.</li><li>Unified utility added to handle nullifier, undefined, and empty values consistently.</li></ul><h4><strong>Transgate Mobile</strong></h4><ul><li>iOS and Android apps completed, with full zkPass protocol integration and optimized performance.</li><li>iSO App Clips and Android Instant Apps launched for quick verification access.</li><li>Integrated all apps with Dev Center for easier business user integration.</li><li>Full testing completed, ensuring compatibility and stability across platforms.</li></ul><h4><strong>Dev Center Enhancements</strong></h4><ul><li>Centralized theme management with modular SCSS for dynamic multi-theme support.</li><li>Optimized rendering performance, reducing latency.</li><li>Containerized with Docker for consistent CI/CD and sub-200ms API response times.</li><li>Redesigned Schema Validator UI with reusable components.</li><li>Refactored message-passing logic for improved reliability and resource safety.</li></ul><h4><strong>VOLEitH Implementation</strong></h4><ul><li>Documentation completed, providing technical guides and API references.</li><li>Development of decentralized verification nodes with multi-signature security.</li><li>Fixed security vulnerabilities in Fiat-Shamir implementation.</li><li>Started research on Circom compatibility for digital circuit integration.</li><li>Developing a recursive Groth16 verifier for efficient proof aggregation.</li></ul><h4><strong>Operations and Maintenance</strong></h4><ul><li>Weekly system checks to ensure stability and integrity.</li><li>Bugs quickly identified and fixed to maintain reliability.</li><li>Strengthened security to protect user data.</li><li>Continuously optimized system performance.</li><li>Actively addressed community feedback to enhance user experience.</li><li>Provided technical support to enterprise clients.</li></ul><h4><strong>Research</strong></h4><p>Research focused on linear congruential generators (LCGs), optimizing randomness and reliability for cryptographic use.</p><p><em>zkPass Links:</em></p><p><a href="https://zkpass.org/"><em>Website</em></a><em> | </em><a href="https://twitter.com/zkPass"><em>Twitter</em></a><em> | </em><a href="https://discord.gg/zkpass"><em>Discord</em></a><em> | </em><a href="https://medium.com/zkPass"><em>Medium</em></a><em> | </em><a href="https://github.com/zkPassOfficial"><em>Github</em></a><em> | </em><a href="http://portal.zkpass.org/"><strong><em>Portal</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=33d0674008d7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/zkpass/monthly-technical-updates-nov-2024-33d0674008d7">Monthly Technical Updates [Nov 2024]</a> was originally published in <a href="https://medium.com/zkpass">zkPass</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>