Center for AI Safety

The Center for AI Safety describes its mission as "to reduce societal-scale risks from artificial intelligence" through research, field-building and advocacy, and runs a compute cluster that it says supports about 20 research labs.¹ Open Philanthropy recommended grants to it of 5,160,000 dollars for general support in November 2022, 4,025,729 dollars for general support, 1,433,000 dollars for the CAIS Philosophy Fellowship, and a final "Exit Grant" of 1,866,559 dollars in October 2023.²

The Extinction Statement

On May 30, 2023, CAIS published a one-sentence "Statement on AI Risk": "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war." Its signatories included Geoffrey Hinton, Yoshua Bengio, Demis Hassabis of Google DeepMind, Sam Altman of OpenAI and Dario Amodei of Anthropic.¹ The Politico reporter Brendan Bordelon noted in October 2023 that the letter had been organized by a group that "received a $5.2 million grant from Open Philanthropy late last year."³

The Action Fund

The Center for AI Safety Action Fund is a 501(c)(4), so contributions to it are not tax-deductible, and it describes itself as "dedicated to advancing public policies that maintain U.S. leadership in AI and protect against AI-related national security threats."⁴ Its report on 2024, its first year, stated that it "led a bipartisan congressional effort" with Representatives Haley Stevens and Jay Obernolte and Senators Martin Heinrich and Mike Rounds that "secured $10 million to establish" the National Institute of Standards and Technology's U.S. AI Safety Institute; that it "co-sponsored California Senate Bill 1047," which the legislature passed and Governor Gavin Newsom vetoed; that it advised the White House and the Commerce Department on the "AI Diffusion" export-control rule; and that it worked with the office of Senate Majority Leader Chuck Schumer on the SAFE Innovation Framework. A photograph of its Capitol Hill launch event in the report shows Jaan Tallinn, co-founder of Skype.⁵ Heinrich and Rounds each had an AI fellow paid through the Open Philanthropy-funded Horizon Institute for Public Service.³ In April 2024 Gina Raimondo named Paul Christiano, a former technical adviser to Open Philanthropy and an original trustee of Anthropic's Long-Term Benefit Trust, head of AI safety at the institute.⁶

Dan Hendrycks

Hendrycks, the center's executive and research director, with Oliver Zhang as managing director,¹ advises Elon Musk's xAI, which he joined in 2023 for a salary of one dollar a year, and Scale AI, which he began advising in 2024 for twelve dollars a year. He told an audience at TechCrunch Disrupt in October 2025 that he had helped craft SB 1047 and that Musk was the only AI company head to support it: "No other A.I. companies officially supported that bill, and that's because he takes this much more seriously."⁷ In July 2024, as SB 1047 moved through the legislature, the AI security company Gray Swan AI announced that Hendrycks, one of its advisers, had "decided to divest his entire financial stake in Gray Swan" "to avoid even the appearance of a financial conflict," and would remain an unpaid adviser without equity.⁸ In March 2025 Hendrycks published "Superintelligence Strategy" with the former Google chief executive Eric Schmidt and Scale AI's founder Alexandr Wang, proposing "Mutual Assured AI Malfunction (MAIM): a deterrence regime resembling nuclear mutual assured destruction (MAD) where any state's aggressive bid for unilateral AI dominance is met with preventive sabotage by rivals," through means "ranging from covert cyberattacks to potential kinetic strikes on datacenters."⁹