Digital Ethics Summit 2026

Antony Walker, Deputy CEO of techUK will open the summit

Keynote

For ten years, techUK’s Digital Ethics Summit has brought together the practitioners, policymakers, technologists, and thinkers at the frontier of advanced data and AI innovation to discuss the potential and power of AI and how this technology can be developed responsibly and in a way that prevents harm. The events of the summer brought the starkest demonstration yet of both the potential and the power of advanced AI. Whether what we have experienced was a watershed moment or a wake up call what is clear is that this conversation has never been more relevant and important. We will open the Summit by leaning into and unpacking the most pressing and important questions and issues of today.

As well as a candid retrospectives of what's worked and what has not over the last 10 years, we will unpack the headlines around the current global race for AI power, control and dominance both politically and commercially and ask what exactly does all this mean for our future.

In light of the race to build and the geopolitical moves to control access to the most powerful AI models, and the concerns being raised by about security, safety and human agency we will ask what this means for the digital ethics discussion and community as well as for the UK in what is a global situation. Leading experts will debate the realities and challenges of the current geopolitical environment where we have started to see examples of power being wielded and the question of AI sovereignty is being asked around the world and consider what the UK, and the international community, can and should now do given what may lie ahead.

Networking

As AI capabilities accelerate, trust is emerging as the defining currency of adoption — the foundation that determines whether institutions, publics, and markets embrace AI or resist it. It's an odd thing to call a currency, and yet "trust" is suddenly everywhere: in policy papers, product launches, and boardroom conversations alike, often invoked more than it's defined. This panel examines what it means to lead on trust, with a particular focus on the UK's approach to safe and secure AI development and governance. We'll look at the geopolitical will shaping this moment, from the UK's early convening role to its broader positioning as a hub for AI safety research, testing, and international cooperation, as well as the forthcoming Geneva AI Summit. We'll also explore the anticipated bill and regulation now taking shape — what frameworks are emerging, how the UK's approach compares to other jurisdictions, and what it takes to build durable trust through governance rather than speed alone. Ultimately, this is a conversation about what responsible leadership in AI looks like in practice, and why trust — not just capability — may be the defining advantage of the years ahead.

Central to the conversation is what it would take for the UK to become the best place in the world to test, evaluate, and assure advanced AI products and infrastructure — and by doing so, attract world-leading developers and innovators to build and deploy here. Given the geopolitical paradigm in which the AI Safety Institute operates, does the UK have an opportunity to move faster than international institutions can, and does it already have what it needs to seize that moment? If not, what action is required, where, and by whom? And underpinning it all: is regulating frontier models the right lever to pull, or is the UK's advantage better secured through other means voluntary standards, testing infrastructure, talent, and investment.

As AI spreads across society, its externalities, both anticipated and unforeseen, are rippling into how we work, live, and relate to one another. Few of these effects are purely good or purely bad, and reasonable people read the same evidence very differently. This discussion brings together contrasting perspectives to examine both sides of the story, with speakers invited to make the strongest case for their view, challenge one another constructively, and find where common ground lies. On the workforce, is AI a catalyst for more meaningful work, new roles, and higher productivity, or a driver of displacement, deskilling, and widening power imbalances at work? On the environment, does AI's promise as a tool for climate modelling, efficiency, and sustainability outweigh the growing energy and resource demands of training and running large models, or are we underestimating the cost? And on the personal and interpersonal, is AI augmenting how we think, connect, and make decisions, or quietly eroding cognition, trust, and wellbeing in an increasingly AI-mediated world? Rather than declaring a winner, this session aims to surface the genuine trade-offs, test the assumptions on each side, and leave the audience better equipped to weigh them for themselves.

Networking

A functioning AI ethics ecosystem doesn't rest on any single actor — it depends on a web of institutions, each playing a distinct role, and on whether those roles add up to a coherent whole. This session takes stock of the institutional landscape as it stands today: who exists, what they're responsible for, and how well those responsibilities interlock in practice. We'll map the current terrain — spanning regulators and government bodies, standards and professional organizations, assurance providers, academic and research institutes, civil society groups, and the newer bodies emerging to fill specific gaps — and ask whether mandates are clear, capacity is sufficient, and coordination is working as intended. Where are the genuine gaps: functions nobody is yet responsible for, capabilities that haven't been built, or mandates that overlap and create confusion rather than coverage? And looking ahead, what new institutions, instruments, or forms of coordination might be needed to close those gaps, and who should be responsible for building them?

Building on this morning's discussion, this session hands the floor to four practitioners for a series of short, sharp lightning talks on how AI assurance actually gets done. Each speaker will share a view from the coal face, showing what it takes to move from principles to practice in environments where the stakes are highest. A year on from DSIT's Trusted Third-Party AI Assurance Roadmap, and with the AI Assurance Consortium (chaired by BCS and supported by techUK's Industry Advisory Group) now moving from ambition to delivery on a voluntary code of ethics and a skills and competencies framework, the talks will take stock of what's been built and test it against what practitioners are seeing on the ground. Speakers will cover the practical trade-offs between in-house audit and third-party assurance as the market professionalises, the progress and barriers in critical national infrastructure and cybersecurity, where practical and meaningful assurance isn't optional, and a deeper shift in thinking from model assurance to mission assurance, asking whether the field has focused too narrowly on individual models and not enough on the systems, infrastructure, and consequences they sit within.

Across the UK, public bodies are exploring how AI can improve services, support staff, and deliver better outcomes for citizens, while making sure it is adopted safely, fairly, and in ways that earn public trust. This breakout session looks at what responsible innovation means in practice for the public sector, and how organisations can move from ambition to real-world impact. The conversation will draw on perspectives from England's regions and the devolved nations, where local authorities, health boards, and other public bodies share common challenges such as skills gaps, budget pressures, and fast-evolving guidance, but operate within different governance structures and at different stages of digital maturity. Panellists will touch on the practical levers that shape adoption, including procurement and how it can enable innovation and local economic benefit, and open the door to smaller suppliers and regional AI ecosystems. We'll also consider how better knowledge-sharing and coordination across nations and regions could help lessons travel further, so that responsible AI adoption becomes something the whole public sector can build on together.

A large-scale, unconference-style participatory session on futures & foresight, held with the goal of getting everyone in the room actively involved and demonstrating, rather than just describing, how we can build and reimagine possible futures together.

Grounding this in historical civil right movements and existing routes to recourse matters as much as looking forward: past moments of technological change can offer hard-won strategic lessons about who gets a seat at the table, how power shifts, and what participatory design can (and can't) achieve. The session will draw on these precedents to sharpen, the scenario work.

Networking

As we close out ten years of the Digital Ethics Summit, this final panel turns from taking stock to looking forward: given everything discussed today, where should the digital ethics community focus its limited time, attention, and resources next? Over the past decade we've seen real progress — principles and frameworks codified, institutions like the Ada Lovelace Institute, the AI Security Institute, and the AI Assurance Consortium established — but we've also seen bodies and initiatives fall away, leaving gaps at precisely the moment the debate feels most urgent. Part of that reckoning means being honest about what has pulled focus rather than sharpened it: where has energy gone into debate, process, or initiatives that haven't moved the needle, and how does the community stay disciplined about what actually matters as the agenda gets more crowded?

This panel brings together the threads raised throughout the day to ask what's genuinely working, what needs to be defended or rebuilt, and what new capability gaps are opening up as agentic AI, scaled deployment, and questions of sovereignty reshape the landscape. It will also look beyond AI to the wider frontier — the convergence of agentic AI, robotics, quantum computing, and immersive and spatial technologies — and ask what the community needs to map now, before deployment outpaces reflection. Rather than another survey of the problem, this is a deliberately practical conversation about prioritization: with finite resources, political will, and organizational capacity, what should the next chapter of digital ethics actually focus on building?