Public Accountability, AI Liability & Regulation with Sandra Wachter | AI Policy100

Professor Sandra Wachter

Professor of Technology and Regulation at the University of Oxford’s Oxford Internet Institute and Humboldt Professor of Technology and Regulation at the Hasso Plattner Institute

October 28, 2026 | 40 min

Who should benefit from the AI revolution, and who should be held accountable when AI causes harm?

In this episode of the Regulating AI Podcast, host Sanjay Puri speaks with Professor Sandra Wachter, Professor of Technology and Regulation at the University of Oxford’s Oxford Internet Institute and Humboldt Professor of Technology and Regulation at the Hasso Plattner Institute, about the future of AI regulation, public accountability, AI liability, and the need to ensure that artificial intelligence serves the public good.

Sandra argues that the public is already a shareholder in AI in all but name. Through public funding, tax incentives, data, electricity, water, and the land used to build data centers, society helps finance the AI industry. So what should the public receive in return?

Drawing on her research, including her paper “Optimizing for Happiness in EU Tech Policy to Repay Public Subsidies for AI,” Sandra makes the case for a new approach to AI governance: one that goes beyond minimizing risks and actively ensures that AI improves people’s lives, strengthens communities, supports education, and benefits society.

The conversation also examines the European Union’s AI regulatory framework, the challenges surrounding AI liability, and the growing debate over whether AI companies should be held responsible when their autonomous agents cause harm. Sandra explains why voluntary commitments and industry promises are not enough, and why enforceable laws are essential to protecting human rights and ensuring accountability.

From AI sovereignty and data center development to the EU AI Act and the limits of existing liability frameworks, this episode explores what it will take to build AI that is not only innovative, but also safe, accountable, and socially beneficial.

CHAPTERS

00:00 Who Should Benefit from AI? The Public’s Stake in Data Centers

01:02 Meet Professor Sandra Wachter

02:00 Why the Public Is a Shareholder in AI

04:48 Holding AI Companies Accountable for Public Subsidies

07:16 Can AI Policy Measure Happiness and Well-Being?

09:44 Beyond Risk: Building AI That Benefits Society

10:47 AI Sovereignty and Enforcing Existing Laws

13:43 The EU AI Liability Directive: What Went Wrong?

16:41 Rethinking AI Harm: Discrimination, Privacy and Misinformation

17:06 Who Should Be Held Liable When AI Causes Harm?

19:39 Europe’s Market Power and the Cost of Saying No

21:17 Does AI Regulation Stifle Innovation?

23:46 The EU AI Act: Transparency, Bias Testing and Human Oversight

25:07 Who Is Responsible When AI Agents Cause Harm?

27:44 Reversing the Burden of Proof in AI Liability

30:18 The AI Act’s Research Exemption and Testing Loopholes

31:50 Should Generative AI Be Classified as High Risk?

32:45 AI Safety Pledges vs. Enforceable Laws

34:49 Why It’s Time for Stronger AI Regulation

36:33 Lightning Round: The Future of AI Governance

38:14 Final Thoughts: AI Must Serve the Public Good

38:41 Sandra Wachter’s New Paper on Happiness and AI Policy

ABOUT OUR GUEST

Professor Sandra Wachter is Professor of Technology and Regulation at the University of Oxford’s Oxford Internet Institute, where she leads the Governance of Emerging Technologies programme. She is also Humboldt Professor of Technology and Regulation at the Hasso Plattner Institute. Her research focuses on the legal, ethical, and societal implications of emerging technologies, including artificial intelligence, algorithmic decision-making, data protection, and fairness.

ABOUT THE REGULATING AI PODCAST

The Regulating AI Podcast, hosted by Sanjay Puri, brings together policymakers, researchers, technology leaders, and experts to examine the laws, institutions, and governance frameworks shaping artificial intelligence. The podcast explores the most pressing questions in AI policy, regulation, accountability, and responsible innovation.

Subscribe to the Regulating AI Podcast for conversations with the people shaping the future of AI governance.

If you found this conversation valuable, share it with colleagues working in AI policy, technology, law, and governance. Let us know in the comments: Should AI companies be legally required to demonstrate measurable benefits to the public in exchange for public support?

#SandraWachter #AIRegulation #AIGovernance #EUAIAct #AILiability #ArtificialIntelligence #ResponsibleAI #TechPolicy #RegulatingAI #SanjayPuri