Press Coverage of the Center for AI Safety Action Fund (CAIS AF)

There are more than 120 AI bills in Congress right now.

The AI industry uses a light lobbying touch to educate Congress from a corporate perspective

Elon Musk Adviser: Are We ‘Sleepwalking’ into an AI TAKEOVER?

AI Congressional Action Threatened by Supreme Court Chevron Blow

An Up-and-Coming AI Safety Thinker on Why You Should Still Be Worried

States Take Up A.I. Regulation Amid Federal Standstill

The Expanding Partnerships For Innovation And Competitiveness (Epic) Act

The AI arms race may soon center on a competition for ‘expert’ data

Researchers Develop New Technique to Wipe Dangerous Knowledge From AI Systems

In Big Tech’s backyard, California lawmaker unveils landmark AI bill

AI Experts Call For Policy Action to Avoid Extreme Risks

Researchers shed light on how to read, control AI systems' minds

Dan Hendrycks, Executive Director, Center for AI Safety

Researchers Poke Holes in Safety Controls of ChatGPT and Other Chatbots

Dan Hendrycks wants to save us from an AI catastrophe. He’s not sure he’ll succeed.

A.I. Poses ‘Risk of Extinction,’ Industry Leaders Warn

Action Fund Statements

View statements from CAIS AF.

Statement on the Center for AI Safety’s Superintelligence Strategy paper.

The Center for AI Safety Action Fund (CAIS AF) Executive Director Varun Krovi released the following statement on the Center for AI Safety’s Superintelligence Strategy paper.

“The Center for AI Safety’s Superintelligence Strategy paper is not a theoretical exercise; it’s a call to action,” CAIS AF Executive Director Varun Krovi said. “The rapid progress of AI is transforming the global national security environment. Superintelligent AI holds the potential to fundamentally alter global balances of power, redefining global interaction, competition, and cooperation among nations.

“The recommendations outlined in the Superintelligence Strategy — centered on deterrence, nonproliferation, and competition — represent the boldest, smartest, and most implementable approach yet to national security in the age of AI,” Krovi said. “To realize the promise of AI, to ensure U.S. leadership, and to guarantee a beneficial global future, policymakers must put this strategy into action.”

This expert analysis, co-authored by Dan Hendrycks, Eric Schmidt, and Alexandr Wang, examines how AI is reshaping global power dynamics, identifies the potential risks posed by rogue actors, and describes the challenges of controlling increasingly autonomous AI systems.

The paper draws on the history of national security policy to offer three core strategies for safe superintelligence.

Deterrence: Introduces the concept of Mutual Assured AI Malfunction (MAIM), a deterrence strategy akin to nuclear Mutually Assured Destruction (MAD), where any state’s aggressive pursuit of AI dominance is met with the threat of preventive sabotage.
Nonproliferation: Emphasizes compute security, information security, and AI security to prevent the proliferation of weaponizable AI capabilities.
Competitiveness: Prioritizes investments in domestic AI chip manufacturing and establishing legal frameworks for AI agents to ensure that the United States remains a leader in this critical technology.

More information about the Superintelligence Strategy paper is available here.