Anthropic CEO Dario Amodei recently proposed slowing down the development of advanced AI models, citing potential safety risks within the next 6 to 12 months. OpenAI CEO Sam Altman and xAI CEO Elon Musk have expressed support for this initiative, which includes third-party safety audits and industry-wide coordination.
Investor and advisor David Sacks criticized the proposal, arguing that if companies believe their models are dangerous, they should simply slow down their own research independently. Sacks warned that seeking government-sanctioned coordination or antitrust exemptions could lead to the formation of an industry cartel that stifles competition.
Sacks advocates for relying on existing product liability laws rather than establishing a new government approval system, such as an AI-specific FDA. He argues that legal and economic consequences for AI-related damages would provide sufficient incentive for companies to manage risks effectively.
The debate highlights a 'prisoner's dilemma' where companies fear that unilateral deceleration would result in a loss of market share and technical advantage to competitors. OpenAI has reportedly sought clarification from U.S. lawmakers on whether industry coordination to slow development would violate antitrust regulations.
U.S. government officials, including President Trump, have expressed concerns that slowing down AI development could jeopardize the country's technological lead over competitors like China. Amodei acknowledged this risk, suggesting that international cooperation is necessary to maintain a balance between safety and competitive advantage.
