What's Next for AI Safety and Regulation?

Former Anthropic and OpenAI researcher Jacob Coxon's viral resignation this month resurfaced a question that the EU AI Act was supposed to settle: Can governments manage frontier AI risk? Coxon's warning that labs are racing toward self-improving superintelligence without adequate safeguards lands just as the EU AI Act's high-risk provisions phase in, and just as critics argue that the legislation regulates compliance paperwork, not the risks Coxon describes.

This panel brings together three perspectives on AI safety and the approaches to regulation in the United States, Europe, and China.  It will look at the various risks and opportunities of AI, and options for addressing challenges at a pace that can keep up with technological development. The conversation will draw on Anthropic CEO Dario Amodei's recent proposal for a three-step plan for developing frontier AI at a more balanced pace to ensure the safety and better alignment of new models. The conversation will close by exploring a practical question: What should EU and US policymakers do next?