Public trust in institutions during tech shifts

When Supreme Court presidents start talking about AI as a “public trust” issue, it is time for the profession to listen very carefully. Lord Reed’s warning in Legal Cheek is not really about technology. It is about legitimacy. He is making three moves that matter for anyone working at the intersection of AI and law: First, he draws a bright line between using AI in the justice system and delegating judicial decisions to it. Hard cases are not binary logic problems; they are value choices in messy fact patterns. If you lost a child contact case and were told “the AI decided”, would you trust the outcome? Most people would not – and trust, not efficiency, is the operating system of the courts. Second, he links AI risks directly to existing trust shocks: populism, politicised attacks on “unelected judges”, and the erosion of respect for independent courts. In that context, dropping opaque, privately owned systems (clustered in one jurisdiction) into the heart of adjudication is not a neutral innovation, it is an accelerant. Third, he points to something technologists often underplay: systems that are trained to please, not to tell hard truths. An AI that optimises for sounding helpful and agreeable will, over time, become more persuasive and less reliable – exactly the opposite of what you want in a forum that settles rights, liberty and family life. For legal and policy leaders, a few implications follow. We should treat AI in courts first and foremost as infrastructure to support human judgment: research, translation, triage, scheduling, accessibility. The threshold for anything closer to “decision‑making” needs to be extraordinarily high, with transparency, contestability and clear lines of accountability baked in. We need to separate “AI for speed” from “AI for legitimacy”. Clearing backlogs with AI tools is attractive and in some contexts entirely appropriate. But if the price is that litigants walk away believing “the computer decided my fate”, the institutional damage will dwarf the efficiency gains. And we should be honest that governance here is geopolitical as well as technical. When a small number of foreign, private actors own the systems that might sit in our judicial back‑end, questions of regulation, security, data control and democratic oversight are not optional extras. AI in the justice system is coming. The challenge – as Lord Reed frames it – is to integrate it in ways that strengthen public trust in courts, rather than quietly hollowing it out in the name of progress.

When it comes to AI, it’s not just about meeting people where they’re at - it’s about meeting people where they trust. I’ve been thinking a lot about scaling AI adoption recently. We often focus on building trust in the tools - making them transparent, explainable, ethical. That matters, but it’s only half the story. The other half is trust in the channel that brings those tools to people. Because trust is transitive. If I trust you, I’m more likely to trust the tool you introduce me to. Global surveys (for example, an Ipsos Public Trust in AI study from 2024) show that while most people are aware of AI, only about one in four feel confident they understand it. Crucially, willingness to use AI rises when it’s introduced by trusted intermediaries - familiar institutions, peers, or community bodies. Academic research backs this up. A 2025 Journal of Management Studies paper found that employees’ engagement with AI depends not just on cognitive trust (does the system perform well?) but on relational or emotional trust - how much they trust the people and institutions implementing it. So perhaps instead of only “meeting people where they’re at”, we should focus on meeting people where they trust - acknowledging the cultural and socioeconomic nuances in how trust shows up in different places. For organisations and governments looking to drive AI adoption at scale, that means: 1. Map trust networks - identify who carries legitimacy: internal leaders, peer networks, professional associations, or local partners / community leaders. 2. Work through trusted channels - adoption rises when AI tools are introduced by credible messengers, not abstract programmes. 3. Build trust not just in the tools but also around the tools - invest as much in relational trust (clarity, co-creation, transparency) as in technical assurance. Because people don’t adopt AI just because it’s good. They adopt it because it’s brought to them by someone they trust. Sources - https://lnkd.in/e3PHgxiu https://lnkd.in/eWZgDB4E #AIAdoption #TrustInAI #ChangeManagement #Leadership #DigitalTransformation #AIforGood Deanna Emeny Carolyn Dawson Christopher Lane Gaurav Gujral Leo Geddes Joe Hildebrand Anna Seely Nick Tate Shirley George Anamaria Duduta Audrey O'Mahony Samuel Holmes Emily Campbell-Ratcliffe

"Fathom’s latest report, AI at the Crossroads: Public Sentiment and Policy Solutions... draws on two national surveys, includes data from 30 focus groups and qualitative interviews, and features takeaways from 150 conversations with leaders across AI, the business community, and civic groups... Four Key Findings ­ 💡 The public is more aware of AI than other key federal issues, but is unsure about what it will mean for them. Over 77% of voters are aware of AI, but they are unsure about its societal impact. Voters are equally excited and concerned about AI’s potential. - Voters display confusion when deciding how much they would trust AI to perform daily tasks or take actions on their behalf. - Generally, voters are more comfortable with AI assisting with tasks than making decisions. - For example, 66% are comfortable with AI analyzing data for schoolwork, but only 26% are comfortable with AI making a purchase using their credit card. - Moreover, 81% are concerned about AI making decisions without human oversight, and they believe accountability and safety measures are essential. The biggest concerns arise when AI could make life-or-death decisions: - 82% of voters are concerned about AI making combat decisions. - 80% are concerned about AI performing surgeries or other medical procedures. 💡 The public wants to balance innovation with the creation of key guardrails, and their priorities do not fall along party lines. Misinformation, deepfakes, privacy, and AI decision-making without human oversight are top concerns for voters. Voters overwhelmingly support specific guardrails to address these issues, including: - Preventing AI interference in elections (84% support) - Ensuring human oversight (84% support) - Protecting data privacy (83% support) - Combating misinformation (82% support) Doomsday scenarios are viewed as alarmist and not compelling. The public breaks from typical partisan tendencies on AI, with: - Democrats doubting government efficacy. - Republicans acknowledging a role for regulation. 💡 The public is concerned about how the tech sector and government are advancing AI. ... - Public trust in the tech sector is waning. Voters are split in their confidence in the companies developing these technologies and worry they will prioritize profit and speed over safety - A strong majority (68%) believe the government has a role in regulating AI, but over half (56%) do not trust the government to regulate it properlyc 💡 The public is looking for a new model of leadership — we need to build a bigger table. ... - Academics and ethicists are favored for their commitment to safety over corporate interests, while elected officials with expertise in technology are seen as crucial for implementing effective regulations - Additionally, the general population should have input, along with professionals from various affected industries, to create a comprehensive and balanced approach to AI governance."

🧠 AI in Government: It’s Not Just Tech — It’s a Transformation of People, Purpose & Trust The call is clear: governments must adopt AI — fast — to improve mission outcomes, public services, and operational effectiveness. But success won’t come from technology alone. It comes from leadership, culture, strategy, and trust. From our just released "Reimagining Government: Achieving the Promise of AI", here are the core principles every public-sector leader should internalize: → AI transformation is first a human challenge. Technical tools matter — but culture, skills, and mindset matter more. Invest equally in people as you do in tech. → Think in portfolios, not isolated projects. Strategic, integrated planning helps avoid redundancy, accelerate impact, and align efforts with mission goals. → Balance innovation & accountability. Innovation without risk awareness can harm trust — but risk controls without innovation stifle progress. Successful advocates manage both. → Partnerships are essential. No agency can go it alone — internal collaboration, external vendors, and human-AI teaming are all foundational. → Leadership must evolve. Emerging roles like Chief Innovation & Transformation Officers bridge mission alignment, culture shifts, tech oversight, and operational excellence. → Use maturity models wisely — but don’t let them slow you. Understand your baseline, but accelerate where feasible to lead, not lag. 📌 Bottom Line: AI isn’t about replacing people — it’s about empowering public-servants, enhancing decision-making, and strengthening public trust. The agencies that succeed will be the ones that treat AI as transformation work, not just procurement work.  Read the full article here: https://lnkd.in/eYiBZ4mF. --- "Reimagining Government" is published by Post Hill Press, distributed by Simon & Schuster, and now available for preorder from all book retailers - Amazon, Barnes & Noble, Inc., Hudson Booksellers, and many others. → Find your retailer (print, e-book, or audio-version) here - https://lnkd.in/ehk2WrCn. → For related research papers, articles & coverage (on Fast Company, Harvard Business Review, MIT Sloan Management Review, etc.), and podcasts, visit - https://lnkd.in/emDzGbYA. ➤ ALL PROCEEDS FROM OUR BOOKS ARE PLEDGED TO CANCER RESEARCH.

If your red line moves when pressure rises, it was never a red line. That’s the tension I keep coming back to watching the AI world this week. We’ve seen companies publish safety pledges. We’ve seen bold language about restraint. We’ve seen public commitments around responsible scaling. And then the competitive pressure increases. Geopolitical stakes rise. Government relationships matter. Market share matters. And the language shifts. Not dramatically. Subtly. But enough. That’s not a tech update. That’s a trust moment. ⸻ What’s actually happening here? We are entering the phase where AI is no longer a tool story. It’s an infrastructure story. Infrastructure shapes information. Information shapes perception. Perception shapes behavior. When companies operating at that level recalibrate safety based on external pressure, the public doesn’t debate the nuance. They internalize one thing. Values are conditional. ⸻ Why does this matter beyond one company? Because we are already in a trust recession. Institutions are fragile. Media credibility is contested. Corporate promises are scrutinized. In that environment, consistency becomes currency. If your safety standards hold only when convenient, trust erodes quietly. Not instantly. But steadily. And steady erosion is more dangerous than scandal. ⸻ Isn’t this just how competition works? Yes. Markets reward speed. But leadership is revealed when speed costs you something. If a company’s red line moves the moment a competitor ships, that tells founders and operators everywhere something subtle but powerful: Performance outranks principle. And that’s the lesson people absorb. ⸻ What does this mean for leaders building with AI right now? It means the question has shifted. We are no longer asking: Can we build this? We are asking: Who is accountable when it scales? If your non-negotiables change based on headlines, capital pressure, or geopolitical incentives, you don’t have guardrails. You have positioning. And positioning doesn’t survive scrutiny. ⸻ Trust in an AI world will not be built by the fastest companies. It will be built by the most consistent ones. The ones whose red lines cost them something. So I’ll ask it plainly: If your competitor crosses a line tomorrow, does that change yours? And if it does… what does that say? Let’s have that conversation. #AI #Leadership #Trust #ResponsibleAI #YOUmanize

The most interesting finding in Microsoft's new AI Diffusion report isn't about technology at all. It's about trust. Singapore, South Korea, and the UAE are outpacing much wealthier nations in AI adoption. Not because they're building the models, but because their governments are actively cultivating public confidence through skilling programs and transparent public sector use. This challenges the assumption that AI leadership naturally flows to those with the biggest compute budgets. Economic advantage is increasingly decoupling from technical capacity and attaching itself to something harder to manufacture: institutional trust and willingness to deploy what exists. The language gap matters too. Multilingual model capability remains a real constraint for ASEAN markets. But countries treating AI adoption as a social and governance challenge, not just a technical one, are creating advantages that capital alone can't replicate. Worth a read if you're thinking about where competitive advantage actually forms in this space. Full report: https://lnkd.in/gVbFQeWb

A new edition of The Long View is out today. This week’s essay explores a question that I believe will increasingly shape leadership, governance, and institutional resilience in the AI era: Can organizations remain trusted while becoming more technologically powerful? As artificial intelligence, automation, and algorithmic systems rapidly reshape institutions, many organizations are simultaneously confronting declining public trust, workforce exhaustion, governance pressures, and growing skepticism around accountability and human oversight. The challenge is no longer simply technological transformation. It is institutional legitimacy. In this edition — “The Trust Imperative: Why Organizational Legitimacy Will Define Leadership in the AI Era” — I reflect on why trust is becoming a form of strategic infrastructure, why governance quality may emerge as a major differentiator between resilient and fragile institutions, and why humane stewardship leadership may become even more important as systems become more automated. The organizations that endure will not necessarily be those that move the fastest. They will be the ones capable of preserving credibility, ethical coherence, human accountability, and public trust during periods of profound transformation. Thank you, as always, to everyone who continues reading and engaging with The Long View. #TheLongView #Leadership #AI #Governance #Trust #InstitutionalLeadership #OrganizationalCulture #Stewardship #FutureOfWork

Two OECD reports just landed over the past two weeks, the latest today. Read together, and with this chart, they tell a story worth paying attention to. The Digital Government Outlook 2026 shows governments are genuinely making progress on AI: strategies in place, guardrails being built, maturity scores up across the board. Then the Trust Survey (2025 wave) reframes everything. Look at this chart carefully. Citizens are cautiously open to AI delivering results: 43% believe it could lead to more tailored services, 42% that it could reduce costs. That's not nothing. But the moment the question shifts to *how* AI is used, confidence drops sharply: only 34% trust government to be transparent about AI use, 35% expect fair and unbiased treatment, and just 32% believe their personal information will be protected. In other words: people can imagine AI being useful. They don't yet trust it to be fair or accountable. This is not a communication problem. It's a delivery problem. Only 6 OECD countries have open algorithm registers. Fewer than half engage citizens when designing AI services. Strategies are strong; implementation lags. The fix isn't a better press release. It's visible accountability. Open registers, genuine user engagement, honest evaluation of what AI is actually delivering. Both reports are out now on oecd.org. Digital Government Outlook 2026 https://lnkd.in/eVxY6Ecm Survey on Drivers of Trust in Public Institutions 2026 Results https://lnkd.in/eFbn6RSs #DigitalGovernment #AI #Trust #PublicAdministration #OECD Jamie Berryhill Tony Tripp Felipe González-Zapata Kenjiro T. Seong Ju Park Julian Olsen Ricardo Zapata Sarah Kups Valerie Frey OECD Public Governance

Everyone talks about AI adoption in government. But the real challenge isn't adoption anymore. It's trust. The UAE has become one of the world's leading examples of digital government and AI-powered public services. From faster service delivery to smarter decision-making, AI is helping transform how public institutions operate and serve communities. Yet technology alone does not determine long-term success. Trust does. Citizens rarely remember how fast a system was. They remember whether it was fair. They remember whether they understood the outcome. And they remember whether help was available when something went wrong. That is where trust is built—or lost. The true test of AI in public services is not what happens when everything works perfectly. It is what happens when an error occurs. Can a decision be explained? Can it be challenged? Is there a clear path for human support when needed? Do citizens understand how outcomes are reached? Transparency, accountability, and human oversight are not barriers to innovation. They are the foundations of public confidence. A UAE-based study on AI adoption in public services identified several key enablers of successful implementation: • Leadership commitment • Clear strategy and governance • Digital infrastructure • Organizational readiness As AI becomes increasingly embedded in government services, perhaps the next stage of maturity is not simply adopting more AI. It is strengthening public trust in the systems already being deployed. Because in government, trust is not a by-product of innovation. Trust is the outcome that matters most. What do you think will be the most important factor in building public trust in AI-powered government services over the next five years? Recent UAE-based research highlights similar findings. Reference: Aljneibi, A. (2020). Enablers & Barriers to AI Adopting in Public Services. The British University in Dubai. #AI #AIGovernance #DigitalGovernment #PublicSector #DigitalTransformation #Leadership #GovernmentInnovation #UAE

It feels like we are approaching an inflection point in how technology is judged. For a long time, most technology decisions were driven by capability and cost. Those criteria still matter. But they are no longer sufficient on their own. Across security, policing, government, and critical industry, I increasingly see a different set of questions emerging alongside them in conversations with leaders responsible for operational risk and long-term resilience. Who builds this technology? What values sit behind it? What incentives shape its development? And what happens if the ecosystem it depends on becomes unavailable, misaligned, or unreliable? This is not about protectionism. It is about resilience and de-risking. Recent years have exposed how dependent many organisations have become on external technology ecosystems for capabilities that sit close to operational decision making and public trust. So technology is now being assessed not only for what it can do, but for whether it aligns with the long-term interests and values of those who rely on it. That shift changes things. It creates space for organisations whose starting point is purpose rather than dominance. For companies that see technology as an enabler of public good, safety. This feels like the next phase of the journey. Capability and innovation will always matter. But increasingly, trust and shared values will matter just as much. In environments where trust is operational rather than theoretical, the character of the technology provider increasingly matters as much as the technology itself.