<?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[Stories by Kizito Witmos on Medium]]></title>
        <description><![CDATA[Stories by Kizito Witmos on Medium]]></description>
        <link>https://medium.com/@kizitowitmos?source=rss-4157b95d1361------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*GLxSXsP0MWxZnGXISMzb-Q.jpeg</url>
            <title>Stories by Kizito Witmos on Medium</title>
            <link>https://medium.com/@kizitowitmos?source=rss-4157b95d1361------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 02:12:37 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kizitowitmos/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[WHEN IT SOUNDS TOO GOOD TO BE TRUE: THE HIDDEN MECHANICS BEHIND MODERN ONLINE MONEY-MAKING SCAMS]]></title>
            <link>https://medium.com/@kizitowitmos/when-it-sounds-too-good-to-be-true-the-hidden-mechanics-behind-modern-online-money-making-scams-430185be59c3?source=rss-4157b95d1361------2</link>
            <guid isPermaLink="false">https://medium.com/p/430185be59c3</guid>
            <category><![CDATA[hacking]]></category>
            <category><![CDATA[africa]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[scams-to-avoid]]></category>
            <category><![CDATA[security]]></category>
            <dc:creator><![CDATA[Kizito Witmos]]></dc:creator>
            <pubDate>Wed, 11 Nov 2026 09:46:14 GMT</pubDate>
            <atom:updated>2026-11-11T09:46:14.457Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6frrIQ-ozhi9aX62vgxpxA.png" /></figure><p>How a $70 sign-up bonus becomes a trap and why your brain keeps falling for it</p><p>You’ve seen it. Maybe a friend shared it in a WhatsApp group, or it popped up in your Facebook feed: “Get paid $70 just for signing up! Complete simple tasks, refer friends, and cash out when you reach $200.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mnZ7hOxK5ABlfkD6zsheZQ.png" /></figure><p>It sounds like easy money. And that’s exactly the point.</p><p>But somewhere between the sign-up bonus and the withdrawal button, something strange happens. You hit $200 and a new pop-up appears: “One more step! Complete Level 3 to unlock your withdrawal.” To reach Level 3, you need to share your referral link in five Facebook groups and wait 24 hours for “verification.”</p><p>The goalposts move. The levels multiply. The withdrawal never comes.</p><p>This isn’t bad luck. It’s a carefully engineered system part psychological manipulation, part technical exploitation. And understanding both sides is the best defense you have.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i32Qgqy639VkgBKnfIrO6Q.png" /></figure><p>The Anatomy of the Loop</p><p>Let’s walk through the lifecycle of these schemes, because the sequence itself is the trick:</p><p>1. The Hook: A promise of easy money ($70 for signing up!) no skills, no investment required.</p><p>2. The Task Wall: Small, achievable tasks get you most of the way to the $200 withdrawal minimum. You’re at $180 and feeling invested.</p><p>3. The Referral Tax: The final stretch requires sharing your link to friends and social groups. Now you’re recruiting others and your reputation is on the line.</p><p>4. The Moving Goalpost: You hit $200. Suddenly there’s a “Level 3” requirement. Then a “verification fee.” Then another level.</p><p>5. The Sunk Cost Trap: You’ve already invested hours and spammed your friends. Quitting now feels like wasting all that effort. So you push further.</p><p>This loop isn’t a bug in the scam, it is the scam. The platform’s real product is your time, your data, your social network, and sometimes your money.</p><p>The Technical Playbook: What’s Actually Happening Under the Hood</p><p>1. Data Harvesting at the Sign-Up Gate</p><p>That “free” account requires an email, phone number, and often a name and country. Attackers aggregate these into bulk databases that are sold on dark web marketplaces or used for phishing campaigns. Your sign-up isn’t the beginning of a payout it’s the product being harvested.</p><p>2. The Moving-Threshold Algorithm</p><p>The “withdrawal minimum” is rarely a fixed number. Scam operators design backend dashboards where thresholds, levels, and conditions can be adjusted in real time. If too many users approach withdrawal, the operators simply add a new requirement. The system is built to keep the percentage of successful withdrawals as close to zero as possible while maintaining the illusion of achievability.</p><p>3. The Crypto Withdrawal Fee Trap</p><p>When the withdrawal method is cryptocurrency, scammers demand an upfront “network fee” or “activation deposit.” This is the cleanest theft in the playbook: crypto transactions are irreversible. Once you send that fee, the money is gone and the withdrawal “fails” or the account is “flagged.” There is no support team. There is no refund.</p><p>4. Malicious Payloads and Fake Apps</p><p>“Download our app to start earning” is a classic vector for malware. Fake APKs and sideloaded apps can install spyware, keyloggers, or banking trojans. Some request excessive permissions contacts, SMS, accessibility services which are then abused to intercept one-time passcodes or spread the scam through your own contact list.</p><p>5. Referral Link Weaponization</p><p>Your unique referral link isn’t just marketing it’s a tracking and trust-laundering tool. When you share it, the scam arrives through a trusted channel (a friend, a family member, a community group). Platforms also use link analytics to map social graphs and measure which users are the most effective recruiters, targeting them for deeper exploitation.</p><p>6. Fake Dashboards and Vanity Metrics</p><p>Modern scam sites display polished dashboards: rising balances, task histories, “pending withdrawals.” None of it is real. It’s a user interface designed to manufacture the feeling of progress. The number on your screen is theater.</p><p>The Psychology: Why Smart People Keep Falling</p><p>If these schemes were obviously fake, no one would fall for them. The trick is that they exploit well-documented, universal cognitive biases.</p><p>The Endowment Effect</p><p>Once your dashboard shows a balance even a fake one you feel like you own it. Psychologists call this the endowment effect: we overvalue what we perceive as ours. Losing “your” $200 feels like a loss, which brings us to the next trap.</p><p>Loss Aversion and the Sunk Cost Fallacy</p><p>People feel losses roughly twice as intensely as equivalent gains. Once you’ve invested time, effort, and social capital, walking away feels like losing everything you’ve built. So you complete one more task, share to one more group, pay one more fee. The scam doesn’t need to convince you to start it just needs to make quitting feel costlier than continuing.</p><p>FOMO and Social Proof</p><p>When a scam arrives via a friend’s referral link, it borrows credibility. “If people I know are doing it, it must be legit.” Add urgency (“This offer expires in 24 hours!”) and scarcity, and the rational brain gets bypassed entirely.</p><p>The Variable Reward Schedule</p><p>Slot machines don’t pay out on a fixed schedule and neither do these platforms. Random small wins, near-misses (“You’re just $20 away!”), and unpredictable progress trigger dopamine responses that make the loop feel exciting rather than suspicious. Behavioral scientists call this variable ratio reinforcement, and it’s the most habit-forming reward structure known.</p><p>The Just-World Bias</p><p>Many victims believe that effort is fairly rewarded that if they just follow the rules and complete the steps, the system will honor its promise. Scammers count on this faith in fairness.</p><p>Red Flags to Memorize</p><p>· A payout that requires an upfront payment (fees, “activation,” “verification deposits”) legitimate employers never charge you to get paid.</p><p>· Withdrawal conditions that change after you qualify.</p><p>· Loyalty built on recruiting friends if your earnings depend on referrals rather than actual work, you’re the product.</p><p>· Apps from outside official app stores, or sites demanding unusual permissions.</p><p>· Urgency and countdown timers designed to prevent you from thinking or researching.</p><p>· No verifiable company identity — no address, no registration, no real support channel.</p><p>How to Protect Yourself</p><p>1. Reverse-engineer the business model. Ask: Who is actually paying me, and why? If there’s no visible product or advertiser, the revenue source is you.</p><p>2. Never pay to get paid. Any withdrawal fee is the scam, full stop.</p><p>3. Google the exact wording. Scam scripts are recycled. Search the platform name plus “scam” or “withdrawal problem.”</p><p>4. Wait 24 hours before sharing anything. Urgency is a weapon; time is your armor.</p><p>5. Verify apps through official stores and check permission requests an “earning app” has no business reading your SMS.</p><p>6. Talk to someone skeptical before committing. A second brain breaks the isolation these schemes depend on.</p><p>The Bottom Line</p><p>Modern scams aren’t crude cons they’re precision-engineered systems that combine dark-pattern UI design, data harvesting, referral manipulation, and a deep understanding of human psychology. The moving goalposts, the 24-hour verifications, the fees, the levels: none of it is accidental. It’s a machine built to convert your hope, your time, and your trust into someone else’s profit.</p><p>So the next time something promises you $70 for doing almost nothing, remember the oldest rule in the book updated for the digital age:</p><p>If it sounds too good to be true, the “too good” part is the bait. The “true” part was never on the table.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=430185be59c3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SECURITY AS FOUNDATION: WHY WE KEEP CHOOSING UNSAFE GROUND]]></title>
            <link>https://medium.com/@kizitowitmos/security-as-foundation-why-we-keep-choosing-unsafe-ground-e03ee34e50e3?source=rss-4157b95d1361------2</link>
            <guid isPermaLink="false">https://medium.com/p/e03ee34e50e3</guid>
            <category><![CDATA[hacking]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Kizito Witmos]]></dc:creator>
            <pubDate>Mon, 09 Nov 2026 08:26:31 GMT</pubDate>
            <atom:updated>2026-11-09T08:26:31.261Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/626/1*xlJX3znvUY3GkuVHVBH9ww.jpeg" /></figure><p>From Cliffside Forts to Unpatched Servers</p><p>For as long as humans have built shelters, we have built them with security in mind. And for almost as long, we have convinced ourselves that the threat wasn’t serious enough to justify the cost.</p><p>A Brief History of Choosing Dangerous Ground</p><p>The earliest settlements tell a story of security as a first-order concern. Hilltop villages, cliff dwellings, and walled towns weren’t aesthetic choices they were risk calculations. A settlement on high ground could see raiders coming. A town with a wall could hold out until help arrived. The location was the security strategy.</p><p>But history is also littered with people who ignored the warning signs. Mesopotamian cities grew prosperous on floodplains despite seasonal floods, because the soil was rich and the trade routes convenient. Roman villas spread into frontier territories with minimal fortification during peaceful decades, right up until they weren’t peaceful anymore. Medieval towns expanded beyond their walls during prosperous times, only to scramble for refuge when conflict returned. And in the modern era, we’ve watched cities grow rapidly in flood-prone deltas and fire-prone wildlands not because residents didn’t know the risks, but because land was cheap, opportunity was close, and the disaster always felt like someone else’s lifetime problem.</p><p>The pattern repeats across millennia: security is prioritized when the threat feels immediate, and quietly deprioritized when growth, convenience, or cost pressure takes over. People don’t usually decide to be unsafe. They decide that safety can wait.</p><p>The Modern Settlement: Your Website</p><p>Now fast-forward to the present. We no longer choose physical ground for most of our ventures, we build on digital ground. A website or application is the modern equivalent of a settlement: it’s where your business lives, where your customers gather, where your reputation and revenue reside.</p><p>And we are making the same calculation our ancestors made.</p><p>Most websites and applications today are built with security as an afterthought, sometimes no thought at all. Features ship first. Deadlines rule. Security reviews get scheduled for “phase two,” which is a lot like promising to build the city wall after the town is already full of people and treasure. Credentials get hardcoded. Dependencies get installed and never updated. Input gets trusted. Access gets over-permissioned. The digital equivalent of building a house with the front door unlocked, the windows open, and a sign out front listing where the valuables are kept.</p><p>The parallels are almost uncomfortably precise:</p><p>Location matters. A house on a floodplain is exposed to floods whether or not the owner thinks about floods. A web application exposed to the internet is exposed to attackers whether or not the developer thinks about attackers. The threat landscape doesn’t require your attention to operate.</p><p>Walls work only if you build them before the raid. Fortifications built during peacetime are cheap; fortifications built mid-siege are impossible. Security controls designed into an application from the start are straightforward; retrofitting them after launch, after the schema is set, the APIs are public, and thousands of users have accounts, is expensive, disruptive, and often incomplete.</p><p>Every unguarded opening is an invitation. A gap in a city wall is not a minor cosmetic issue; it’s where the attack happens. A missing authorization check, an unvalidated input field, a leaked API key, these aren’t minor technical debts. They’re the gap in the wall, and automated scanners will find them the way scouts find an unguarded gate.</p><p>The disaster always feels hypothetical, until it isn’t. The frontier villa owner reasoned that raids were rare. The floodplain merchant reasoned that the last flood was a generation ago. The modern builder reasons that “nobody would target us.” Then the breach happens, and the calculation changes overnight: data exposed, customers lost, trust destroyed, and a frantic scramble to build in months what should have been built in days.</p><p>What a House at Risk Actually Looks Like:</p><p>Picture two houses on the same street.</p><p>The first was built with intent: solid foundation, deadbolts, reinforced frames, good lighting, a monitored alarm, and insurance. The owner spent maybe 5–10% more upfront and sleeps well.</p><p>The second was built fast and cheap. The contractor skipped the foundation inspection because the schedule was tight. The back door doesn’t lock properly but “it’s fine, we’re usually home.” The spare key lives under the mat. The owner saved money and time and has no idea that the neighborhood has recently seen a wave of break-ins.</p><p>Now map that onto two applications. One was designed with authentication, authorization, input validation, encryption, dependency management, and monitoring baked in from the first commit. The other shipped fast, stores passwords in plaintext (or close to it), trusts client-side validation, never patches its libraries, and logs nothing. Both may look identical to a visitor. Only one survives contact with an attacker.</p><p>The Uncomfortable Truth</p><p>Our ancestors eventually learned often the hard way that security isn’t a feature you add to a settlement. It’s a property of how the settlement is designed. The cities that endured weren’t the ones that got lucky; they were the ones that treated defensibility as foundational.</p><p>We have the same lesson available to us, and we don’t have to learn it through a breach. Security in software, like security in settlement, is cheapest and most effective when it’s part of the ground plan, not a wall hastily thrown up after the raiders are already inside.</p><p>The question isn’t whether your website faces risk. It does, the moment it’s online, just as every house faces risk the moment it’s built. The question is whether you’re building like someone who knows it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e03ee34e50e3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[THE UNPLUGGED LOCK: WHY MALAWI NEEDS ITS ETHICAL HACKERS BEFORE THE NEXT BREACH]]></title>
            <link>https://medium.com/@kizitowitmos/the-unplugged-lock-why-malawi-needs-its-ethical-hackers-before-the-next-breach-28ebe2c79c91?source=rss-4157b95d1361------2</link>
            <guid isPermaLink="false">https://medium.com/p/28ebe2c79c91</guid>
            <category><![CDATA[cyber-security-awareness]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[malawi]]></category>
            <category><![CDATA[information-security]]></category>
            <dc:creator><![CDATA[Kizito Witmos]]></dc:creator>
            <pubDate>Sun, 25 Oct 2026 10:48:19 GMT</pubDate>
            <atom:updated>2026-10-25T10:48:19.457Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*6yzXkLlzaSgCUbuVVSbp1g.png" /></figure><p>Malawi is building its digital economy on software it has not checked. Websites are launched, payment portals go live, customer databases fill up, and the servers underneath are left to age in public. The vulnerabilities are not hidden. They are indexed, scored, and published for anyone to read. Yet the people who know how to find them, ethical hackers, penetration testers, self-taught researchers cannot get hired. The industry is too young, the budget lines too thin, the conversation too uncomfortable. So the servers stay online. The locks stay unplugged.</p><p>TNM offers a useful case study, not because it is uniquely careless, but because it is one of the few Malawian organizations that has publicly acknowledged the problem and, at least in one job posting, shown what a solution could look like.</p><p>What TNM Revealed:</p><p>In October 2026, TNM confirmed that customers had reported their WhatsApp accounts being compromised. The company said it had engaged Meta, the platform’s operator, to resolve the issue, which affected users in several countries. TNM advised customers to enable two-step verification, avoid sharing codes, and check connected devices. Crucially, the available information did not establish that TNM’s own systems had been breached .</p><p>That distinction matters. But it also reveals the shape of the problem. A telecom operator can issue a security advisory and point to a foreign platform. It cannot, however, patch the skills gap that leaves its own infrastructure and its customers exposed to the next wave. And the next wave is already here. Reports from Malawi24 quoted users who lost WhatsApp Business accounts despite having two-step verification enabled and email recovery linked . Whether those specific incidents trace back to a network-level issue or a broader credential compromise remains unclear. But the pattern is unmistakable: security is being treated as a customer-education problem rather than an engineering one.</p><p>TNM knows better. In July 2025, it advertised for a Patch and Vulnerability Management Specialist. The requirements were explicit: CompTIA Security+, Certified Ethical Hacker, proficiency with Nessus and Qualys, “understanding of ethical hacking principles” . The job description was not vague. It asked for someone who could perform internal penetration testing, manage Web Application Firewalls, and review vulnerability data to prioritize patching .</p><p>This is exactly the right instinct. But it is also the exception that proves the rule. If one of Malawi’s largest telecommunications companies needs a dedicated specialist with ethical hacking skills to keep its environment safe, what does that say about the hundreds of smaller organizations running websites on unpatched servers with no one on staff who knows what a CVE is?</p><p>The Servers Are Still Running:</p><p>The nginx vulnerability situation is a concrete illustration of the gap between knowledge and action. As of May 2026, Ubuntu’s own bug tracker confirmed that CVE-2026–42945, a heap-based buffer overflow in nginx, affected every version shipped in supported releases, including nginx 1.24.0 on Ubuntu Noble . The flaw is reachable remotely without authentication. It causes deterministic heap corruption. Repeated requests can crash worker processes in a loop. Code execution is possible on builds where ASLR is disabled . The fix was backported and published. The advisory was public.</p><p>And yet nginx 1.24.0 is still widely visible in HTTP headers across Malawian websites. Not because the patch does not exist, but because no one with the skills has been tasked with applying it. No one has been paid to look. The vulnerability is not a secret. The patch is not hidden. The only missing ingredient is a person who knows how to check, and an organization willing to let them.</p><p>This is the certification trap in its purest form. Malawi’s Cybersecurity Act of 2024 includes provisions for “the accreditation of cybersecurity professionals and practitioners” . Accreditation is a reasonable goal. It creates standards and protects against frauds. But in a country where the formal security industry is still nascent, accreditation can become a gate that keeps out the very people who have already been doing the work. The self-taught researcher who found a misconfiguration in a government portal and reported it privately. The developer who learned Burp Suite at night because no employer would pay for a course. The young analyst who can read a CVE advisory and immediately understand what it means for a Malawian bank’s public login page.</p><p>TNM’s job posting asked for CEH certification alongside ethical hacking understanding . That is a reasonable standard for a large employer with a formal security division. But most Malawian organizations do not have a security division. They have IT generalists. They have web developers who were asked to “make it secure” as an afterthought. They have no one who can look at an nginx error log and recognize it as a symptom rather than a nuisance.</p><p>The result is a paradox. There are skilled people in Malawi who can find the holes. There are organizations with holes that need finding. Between them sits a hiring culture that says: come back when you have a certificate, come back when we have a budget line, come back when someone else has already been breached and made it embarrassing enough for us to care.</p><p>What Ethical Hackers Actually Do:</p><p>It is worth being precise about the value proposition, because “ethical hacking” sounds like a contradiction to ears trained on old hierarchies. A penetration tester does not break things for the thrill of it. They break things in a controlled, documented, reproducible way so that the organization can fix them before someone with worse intentions does the same thing without permission.</p><p>When a tester scans a Malawian site and sees a server banner advertising an outdated nginx version, they are not just reading a header. They are checking whether that version is vulnerable to known exploits. They are testing whether the server leaks information through malformed requests. They are looking for the forgotten administrative panel, the default credentials, the file upload endpoint that accepts an SVG with a script inside it. This is not magic. It is method. And it is method that Malawi desperately needs applied to its public-facing infrastructure.</p><p>The WhatsApp compromises that TNM acknowledged were not caused by brilliant criminals exploiting a zero-day. They were caused by the ordinary, persistent failure of systems that were never tested by someone whose job was to think like an attacker. A red team exercise would have simulated exactly that kind of social engineering and credential theft. A security assessment would have identified the weak points. The cost of such an assessment is a fraction of the cost of the breach.</p><p>The Questions They Should Be Asking:</p><p>If you run an organization in Malawi with a website, a payment portal, or a customer database, here are the questions you should be able to answer. If you cannot, the problem is not that you are too small to be targeted. The problem is that you have not looked.</p><p>What version of nginx or Apache is serving your site right now, and when was it last patched? If the answer is “I don’t know” or “the hosting provider handles that,” you are trusting a third party with your security without a contract that says so. The CVEs are public. The scanner is free. There is no excuse for ignorance.</p><p>When did you last have an independent person not your developer, not your IT vendor attempt to break into your systems? If the answer is never, you are relying on the hope that no one has tried. Hope is not a control.</p><p>If an attacker gained access to your customer database tomorrow, how would you know? Detection is not optional. Logging is not optional. The absence of an incident response plan is a plan to panic.</p><p>Would you hire someone who can prove they found a vulnerability, even if they do not have a certificate? If the answer is no, you are prioritizing paperwork over proof. The best security researchers in the world are often self-taught. Malawi cannot afford to import all of its expertise. It must cultivate what it has.</p><p>The Path Forward Is Not Complicated:</p><p>Malawi does not need a Silicon Valley security industry to protect itself. It needs a culture shift. It needs organizations to stop treating cybersecurity as a cost center and start treating it as a precondition for doing business online. It needs a hiring pipeline that recognizes skill in whatever form it arrives certified or not, formal or self-taught and puts that skill to work on the servers that are already exposed.</p><p>TNM’s decision to hire a patch and vulnerability specialist is a step in the right direction. But one company’s security posture does not protect the ecosystem. The Cybersecurity Act of 2024 gives Malawi a framework. MACRA’s mwCERT gives it an institutional home . But institutions do not patch servers. People do. And the people are here, waiting for someone to ask them what they found.</p><p>The WhatsApp accounts are still being compromised. The nginx 1.24.0 servers are still running. The next breach is not a question of if. It is a question of who will be paid to find it first an ethical hacker with a laptop and a conscience, or someone who does not care about Malawi at all.</p><p>The choice is not technical. It is moral. And it is due.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=28ebe2c79c91" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[WHY AFRICA SHOULD NOT STAY SILENT ON PHP2SHELL]]></title>
            <link>https://medium.com/@kizitowitmos/why-africa-should-not-stay-silent-on-php2shell-bbafd01b8f9b?source=rss-4157b95d1361------2</link>
            <guid isPermaLink="false">https://medium.com/p/bbafd01b8f9b</guid>
            <category><![CDATA[malawi]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[africa]]></category>
            <category><![CDATA[php2shell]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <dc:creator><![CDATA[Kizito Witmos]]></dc:creator>
            <pubDate>Fri, 23 Oct 2026 10:41:23 GMT</pubDate>
            <atom:updated>2026-10-23T10:41:23.929Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Q0rji763uz9csuwe5lb7Hg.png" /></figure><p>Most of the internet runs on PHP. We are talking about roughly 70% of all websites. So when a security hole is found inside the very engine that makes PHP work not in a single app, but in the foundation itself every website built on it becomes a possible way in.</p><p>Africa is building its digital economy fast. And the security gaps are real. So this is not a distant problem. It is ours too.</p><p>In October 2026, the cybersecurity world faced something Microsoft CEO Satya Nadella said he had not seen “in a long, long time.” Two critical flaws in PHP were discovered. They were named CVE-2026–109618 and CVE-2026–109619, and together called PHP2Shell. Both received a severity score of 9.2 out of 10.</p><p>What happened next was frightening. Within 19 minutes of the fix being released, attackers were already trying to break in. Cloudflare recorded millions of attempts within hours.</p><p>So for African organizations, the question is simple: Are you ready? Because you are almost certainly affected.</p><p>What Is PHP, and Why Is This Flaw Different?</p><p>PHP stands for “PHP: Hypertext Preprocessor.” It started in 1994 as a simple tool for building web pages. It was free, easy to use, and worked well with databases like MySQL. That is why it became the backbone of the modern web. Today it powers about 70% of all websites. WordPress, Drupal, Joomla, Magento many of the platforms we use every day all rely on PHP.</p><p>What makes PHP2Shell so dangerous is where the hole is. Most PHP problems sit in apps or plugins: a badly written login page, an old WordPress extension. PHP2Shell sits in the PHP engine itself the part that runs every line of code on the server.</p><p>This matters a lot. A perfectly written, fully updated website can still be hacked simply because the engine underneath it is broken. CrowdStrike put it clearly: “There is no authentication to bypass, no session to steal, and no vulnerable endpoint to discover. The exposure is independent of the CMS, framework, or plug-ins in use.”</p><p>In simple terms: attackers do not need to find a weakness in your website. They only need to find your server.</p><p>The attack is also clever and quiet. Public tools like php2shell.py showed how an attacker could run commands on a server without logging in. Some tools even send the results straight back in the web response. You do not need a special tool like nc (netcat) to keep a connection open. The command output just comes back with the page load.</p><p>CrowdStrike warned that exploitation leaves “virtually no digital traces.” No strange app logs. No odd responses. Just a routine memory error that most IT teams ignore as normal noise.</p><p>And here is the twist: this flaw was found using AI. Researchers at Wiz and Calif used advanced AI models, including OpenAI’s GPT-5.6 Sol, to find problems that had been hidden for years too complex for humans to spot practically. AI made finding the bug easier. It also makes it easier for attackers to use it.</p><p>The Global Panic and Where Africa Stands:</p><p>The world reacted fast. The UK’s National Cyber Security Centre, Canada, Australia, and the EU’s CERT-EU told organizations to patch or shut down. Finland’s government took public services offline. The Dutch House of Representatives went offline. The BBC had downtime. Japan’s Asahi and France’s Orange cut off access to PHP-based systems.</p><p>These are rich organizations with full security teams. And they still struggled.</p><p>Now picture the African context.</p><p>INTERPOL’s African Cyberthreat Assessment 2026 found that AI was involved in 55% of reported cybercrimes across the continent. Reported losses more than doubled from $192 million to $484 million since 2024. More than 90% of African businesses lack proper cybersecurity rules. The continent is short of over 100,000 cybersecurity professionals. Many countries still do not have working Computer Emergency Response Teams (CERTs) or digital forensics units.</p><p>When a flaw as big as PHP2Shell appears, these gaps become life-threatening for businesses. An African online shop running an old version of PHP maybe because no one knew it needed updating, or because the hosting company never told them is not a “maybe victim.” It is the likely victim.</p><p>A small business in Lilongwe, Lagos, Nairobi, or Accra that built its website on a PHP-based system has no security operations center watching for a single memory error in the logs. It has no one to detect the attack, let alone respond.</p><p>The quiet nature of the attack makes it worse. Unlike ransomware that announces itself, PHP2Shell can work silently stealing passwords, copying data, planting backdoors while the website keeps running normally. By the time an African organization notices, the damage may be months old.</p><p>Why Silence Is Not an Option:</p><p>Some people argue that Africa’s lower internet use protects it from global cyber events. That is dangerously wrong. Africa’s digital growth is speeding up: internet use grew from 10% in 2014 to 43% in 2024, and is expected to reach 1.1 billion users by 2029. Mobile money, e-government services, healthcare systems, and financial infrastructure are increasingly built on web technology and a large part of that web runs on PHP.</p><p>When an attacker exploits PHP2Shell on a mobile money server, the victims are not abstract “users.” They are real people whose savings disappear, whose loan applications are compromised, whose identities are stolen. INTERPOL found that East Africa has become a hub of mobile money fraud. A flaw that allows silent server takeover is a direct threat to that ecosystem.</p><p>The global response also shows a hard truth: coordinated disclosure depends on trust and capacity. When Chainguard released PHP images with the fix before the agreed disclosure date, the PHP Foundation noted that “any attacker could now study the fix in binary form while every other PHP user was still waiting.” African organizations, already at the back of the patching queue, are the least able to handle such shocks.</p><p>What Africa Must Do:</p><p>First, patch. Immediately. The fixed versions are PHP 8.5.12, 8.4.27, 8.3.36, and 8.2.35. Ubuntu and other systems have released security updates. If you cannot patch, take systems offline. There is no workaround. As CERT.UG stated plainly: “no workaround exists outside of patching.”</p><p>Second, know your attack surface. Many African organizations do not have a full list of software running on their servers. PHP is often installed by a hosting provider or bundled into a CMS, invisible to administrators. If you do not know which PHP version your website runs, you cannot know if you are exposed. This is not a technical luxury it is basic hygiene.</p><p>Third, invest in the unglamorous work. Africa’s cybersecurity gap is not a treaty problem; it is a readiness problem. The Malabo Convention gives a legal framework, but frameworks do not patch servers. What Africa needs is a culture of maintenance: routine updates, vulnerability monitoring, incident response drills, and a trained workforce. INTERPOL’s Operation Serengeti 2.0 showed that coordinated action across 18 African countries and the UK can work 1,209 arrests, infrastructure dismantled, funds recovered. That model of practical cooperation must be used for vulnerability response too.</p><p>Fourth, demand accountability from the global tech ecosystem. PHP is free, open-source software run by a small foundation. When critical flaws appear, the burden falls on users. African organizations, as a collective market, have power to demand better security defaults, clearer communication, and support for old systems from hosting providers and platform vendors.</p><p>The Silence Has a Cost:</p><p>The global cybersecurity community responded to PHP2Shell with urgency because the numbers were huge: millions of websites, millions of exploit attempts, governments going offline. Africa’s numbers are smaller in absolute terms, but the relative impact is greater. A smaller digital economy, thinner security resources, and faster growth mean that one large-scale attack could destroy trust in the continent’s digital future.</p><p>When Satya Nadella said he had not seen anything like this “in a long, long time,” he was speaking as the leader of one of the world’s most resourced tech companies. African organizations do not have the same cushions. But they do have something else: the knowledge that this is coming, and the choice to prepare rather than wait.</p><p>The patch exists. The exploit needs no login. The traces are almost invisible. The only question is whether Africa will act before the attack arrives or explain afterward why it stayed silent.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bbafd01b8f9b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[WHAT PAYCHANGU HAS BUILT AROUND PAYOUT SECURITY]]></title>
            <link>https://medium.com/@kizitowitmos/what-paychangu-has-built-around-payout-security-6315fdf14715?source=rss-4157b95d1361------2</link>
            <guid isPermaLink="false">https://medium.com/p/6315fdf14715</guid>
            <category><![CDATA[payment-gateway]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[malawi]]></category>
            <category><![CDATA[cyber-security-awareness]]></category>
            <category><![CDATA[ethical-hacking]]></category>
            <dc:creator><![CDATA[Kizito Witmos]]></dc:creator>
            <pubDate>Thu, 22 Oct 2026 10:01:51 GMT</pubDate>
            <atom:updated>2026-10-22T10:01:51.224Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*6eTSojCthzu5WB48QstsRw.png" /></figure><p>PayChangu has steadily assembled one of the more thoughtful security stacks in Malawi’s payment space, and its latest dashboard controls for payouts deserve close attention from any organization moving money through the platform. The gateway now requires developers to manually enable payouts and whitelist a server IP before any payout API call will succeed. Handled well, this is a genuine second lock on the vault. Handled carelessly, it offers less protection than it appears to. Understanding both sides of that coin is essential for anyone responsible for disbursements.</p><p>A Meaningful Step Forward:</p><p>The core idea behind the update is sound and welcome. Previously, payout security rested largely on the secrecy of the API secret key. If that key leaked through a compromised server, a careless commit, or an insider, an attacker could initiate payouts to any destination they controlled, and the money would move instantly. PayChangu has now introduced a deliberate two-step activation: payouts stay off until manually enabled in the dashboard, and requests are only accepted from IP addresses the developer has explicitly whitelisted. A stolen key is far less useful if the attacker is not sitting on the approved server. This mirrors the IP allowlisting that mature financial infrastructure providers treat as standard for withdrawal endpoints, and it reflects a platform that is listening to the security concerns of its integrators.</p><p>The Configuration Gap Worth Understanding:</p><p>There is one important nuance that every merchant should appreciate. If payouts are enabled but no IP is whitelisted, or if the whitelist is set too broadly, the secret key once again becomes the only barrier. This is not a criticism of the feature so much as a reminder that configuration-dependent controls are only as strong as the discipline behind them. A developer rushing to ship a payout feature might toggle it on, test it from a local machine, and never return to complete the whitelist properly. The most robust design would see the dashboard decline to enable payouts until at least one IP is registered, and it would be worth confirming with PayChangu whether that guardrail is enforced. In the meantime, the responsibility sits clearly with the integrator, and that is a reasonable place for it to sit.</p><p>Regulatory Legitimacy as a Security Foundation:</p><p>The dashboard controls do not exist in isolation. The Reserve Bank of Malawi granted PayChangu a Payment Service Provider licence in September 2026, moving the platform from a convenient API into regulated infrastructure. That matters for security because licensed PSPs operate under supervisory expectations for fraud controls and transaction monitoring. Organizations choosing a gateway are not simply choosing an endpoint; they are choosing the compliance culture that surrounds it, and PayChangu’s regulatory standing is a meaningful part of that picture.</p><p>Delegated Authorization Through Connect:</p><p>For SaaS platforms and marketplaces, PayChangu Connect represents a thoughtful architectural choice. End users authenticate separately and grant scoped permissions via access tokens, which means the platform never holds the keys to the customer’s wallet directly. The practical effect is a smaller blast radius if a platform is breached. This is the kind of design decision that reduces systemic risk rather than simply shifting it, and it complements the payout controls nicely.</p><p>Webhook Signatures and Server-Side Verification:</p><p>PayChangu’s webhook model is well considered. Every webhook carries a Signature header containing a SHA-256 HMAC hash of the payload, generated with the developer’s web secret key, and mismatched requests are discarded. The documentation also encourages developers to re-query the API after receiving a webhook before releasing value, and to verify incoming payments against the expected amount, currency, and transaction reference. These are sound practices that align with how established gateways around the world handle event integrity. The new payout controls are the natural counterpart for outgoing money.</p><p>A Practical Checklist for Organizations:</p><p>For businesses using PayChangu payouts, a short audit is worthwhile. Confirm that the production server has a static egress IP before whitelisting it. Whitelist only that address rather than a broad range. Test the rejection path in sandbox by sending a payout from a non-whitelisted IP and confirming it fails. Monitor for rejected payout attempts, since a spike is the clearest signal that someone has the key and is trying to use it from elsewhere. Rotate the secret key immediately if exposure is suspected. And keep payouts disabled in production until the whitelist is confirmed working, treating the toggle as a control rather than a formality.</p><p>The Bottom Line:</p><p>PayChangu has done the right thing by giving developers the tools to lock down payouts. The manual toggle and IP whitelist are meaningful improvements over the previous model, and they sit within a broader security posture that includes regulatory licensing, delegated authorization, and signed webhooks. What remains is the human layer: security features that depend on correct configuration are only as strong as the discipline behind them. Organizations that treat the new controls as a checklist to complete rather than a box to tick will find themselves well protected. The gateway has provided a second lock, and the wisest course is to use it before anyone else tries the door.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6315fdf14715" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>