The AI Policy Podcast

This week, Aalok is joined by Maryam Mujica, president of the General Catalyst Institute, and Munjal Shah, co-founder and CEO of generative AI startup Hippocratic AI, for a discussion on how generative AI startups are thinking about AI regulation. Guests:  Maryam Mujica, President, General Catalyst Institute Munjal Shah, co-founder and CEO, Hippocratic AI Timestamps: About the General Catalyst Institute (1:52) About Hippocratic AI (6:56) Healthcare applications of AI (13:43) Impact of regulation on AI startups (18:32) Munjal's interactions with federal regulators (24:07) Issue of compute shortage (31:21) Impact of AI on labor (48:23)

An Anthropic researcher's viral resignation post last week set off a maelstrom of fear around extreme AI risks. Aalok and Nicole discuss the resulting vibe shift and how frontier labs are responding. They also unpack OpenAI solving one of the most difficult open problems in mathematics and reports of more rogue agent activity on the open internet.   Timestamps: OpenAI solves Millenium Problem (1:04) Rogue OpenAI agents take over a German wiki (12:14) Anthropic researcher's resignation post goes viral (19:34) Dario Amodei commits to embedding third-party evaluators inside Anthropic (28:12) Where the U.S. government can step in (35:12) Additional Reading: "On the Navier-Stokes Millennium Prize Problem" by OpenAI: https://openai.com/index/navier-stokes-solution/ "An OpenAI model has disproved a central conjecture in discrete geometry" by OpenAI: https://openai.com/index/model-disproves-discrete-geometry-conjecture/ Statement on Navier-Stokes by NYU Professor Tristan Buckmaster: https://cims.nyu.edu/~tristanb/statement.pdf "An alignment assessment of recent cybersecurity incidents" by Anthropic: https://www.anthropic.com/research/alignment-assessment-cybersecurity-incidents "OpenAI's rogue agents used at least 10 more sites for unauthorized comms, researchers say" by Reuters: https://www.reuters.com/world/openais-rogue-agents-used-least-10-more-sites-unauthorized-comms-rese… "OpenAI agents carried out an undisclosed cyber-attack on RubyGems" by Spencer Kitts, Thomas Larsen, and Sydney Von Arx: https://www.rubyhack.ai "Discovery of a new OpenAI agent message board" by Sydney Von Arx, Cormac Slade Byrd, Spencer Kitts, and Thomas Larsen: https://collusion.wiki "Anthropic Researcher Quits Over 'Out-of-Control' AI Fears" by The Wall Street Journal: https://www.wsj.com/tech/ai/anthropic-researcher-quits-over-out-of-control-ai-fears-707b7628 "The AI policy window is open. We need to act." by OpenAI Chief Global Affairs Officer Chris Lehane: https://openai.com/index/ai-policy-window/ "We Must Pace the Frontier" by Dario Amodei: https://darioamodei.com/post/we-must-pace-the-frontier

This week, Aalok is joined by Jason Elliott, the former Deputy Chief of Staff to California Governor Gavin Newsom, for a conversation about regulating AI at the state level.   Guests: Jason Elliott, President, Versus Solutions   Timestamps: Jason's bio (0:16) Legislating at the state vs. federal level (2:21) Revisiting SB-1047 (5:58) SB-53 (13:43) States as "laboratories of democracy" (19:57) Influence of SB-53 on subsequent state bills (24:14) Federal preemption debate (27:59) Updating existing regulations (40:37)   Additional Reading: CA Senate Bill No. 1047: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202320240SB1047 CA Senate Bill No. 53: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53 "The California Report on Frontier AI Policy" by the Joint California Policy Working Group on AI Frontier Models: https://www.gov.ca.gov/wp-content/uploads/2025/06/June-17-2025-–-The-California-Report-on-Frontier-AI-Policy.pdf "Toward a Federal Framework: Lessons from State and International Frontier AI Regulation" by Laura Caroli and Aalok Mehta: https://www.csis.org/analysis/toward-federal-framework-lessons-state-and-international-frontier-ai-regulation CA Senate Bill No. 813: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB813 "Independent Oversight Marketplace for AI" by Fathom: https://ivo.fathom.or

This week, we unpack findings from METR's six-day investigation of the OpenAI-Hugging Face cyber incident. We also discuss an IVO bill passed by the California Legislature, Bill Gates' take on the current moment in AI, and a federal judge striking down Anthropic's supply chain risk designation. Additional Reading: SB-813 bill text: https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202520260SB813 "Toward a Federal Framework: Lessons from State and International Frontier AI Regulation" by Laura Caroli and Aalok Mehta: https://www.csis.org/analysis/toward-federal-framework-lessons-state-and-international-frontier-ai-regulation "The turbulent AI era is here. The choices we make now are critical." by Bill Gates: https://www.gatesnotes.com/a-turbulent-ai-era-and-critical-choices-to-make "Brief independent investigation of agents' behavior, reasoning and collaboration in the OpenAI/Hugging Face hacking incident" by METR: https://metr.org/blog/2026-08-26-openai-hugging-face-incident-investigation/ "The Hugging Face incident and the road ahead" by OpenAI: https://openai.com/index/hugging-face-incident-and-the-road-ahead/ "A call for collective action on cyber defense" open letter by OpenAI: https://openai.com/collective-cyberdefense/

This episode cross-posts a panel from the event "AI Agent Containment Failures: Technical Realities and Policy Responses," co-hosted by the Wadhwani AI Center and the Institute for Law and AI on August 24.  Check out the full event recording: https://www.csis.org/events/ai-agent-containment-failures-technical-realities-and-policy-responses   Guests: Helen Toner, Executive Director, Center for Security and Emerging Technology (CSET) at Georgetown Mackenzie Arnold, Director of U.S. Policy, Institute for Law and AI (LawAI) Matt Pearl, Director, Strategic Technologies Program at the Center for Strategic and International Studies (CSIS)  Timestamps: What we've learned since the OpenAI-Hugging Face incident (1:15) Incentives for safety measures at frontier labs (16:24) Technical talent within government (22:18) Ensuring compliance to incident reporting requirements (34:58) Liability for crimes committed by AI agents (38:00) Liability safe harbors (51:55)   Additional Reading: "When Reporting an AI Security Incident Is Not Mandatory" by Mackenzie Arnold and Stephan Llerena for Lawfare: https://www.lawfaremedia.org/article/when-reporting-an-ai-security-incident-is-not-mandatory "Investigating three real-world incidents in our cybersecurity evaluations" blog by Anthropic: https://www.anthropic.com/news/investigating-incidents-cybersecurity-evals "Incident Report: unsanctioned agent behavior during cyber testing" by U.K. AISI: https://www.aisi.gov.uk/blog/incident-report-unsanctioned-agent-behaviour-during-cyber-testing "Pacing model development in an era of cyber-critical capabilities" announcement from OpenAI: https://openai.com/index/pacing-model-development-cyber-capabilities/ "Pacing the Frontier" petition: https://www.pacingthefrontier.com/ CSIS Commission on U.S. Cyber Force Generation: https://www.csis.org/analysis/csis-commission-us-cyber-force-generation Limitations of the current incident reporting regime (7:46)Role of the U.S. government in helping defenders (11:52)What the executive branch can do now (25:29)Concrete policy recommendations (29:31)U.S.-China competition (40:50)Preventing abuse of an incident reporting regime (44:25)De facto regulatory role played by frontier lab employees (49:24)Role of AI-generated code in cyberdefense (54:45)Concluding remarks (55:53)"Out of Bounds: What the U.S. Government Should Do in Response to AI Agent Containment Failures" by Aalok Mehta: https://www.csis.org/analysis/out-bounds-what-us-government-should-do-response-ai-agent-containment-failures"The 'Breaking' News: The OpenAI–Hugging Face Incident" presentation at Black Hat: https://youtu.be/87DyyMV0kCY?si=GWh9MB1plhO-xjYT