U.S. President Donald Trump signed a voluntary agreement with executives from major AI companies on managing the safety of advanced AI models. Cointelegraph, a blockchain media outlet, reported on Sept. 30 local time that the deal centers on leaving much of the responsibility for managing frontier AI risks to developers.

The White House agreement is titled "Joint Commitment on Frontier Responsibility". It calls for internal controls, independent external audits and board-level oversight to prevent AI models from accessing or hacking technical systems in unintended ways. It also left open the possibility of turning the voluntary pledges into future laws or regulations.

Trump told reporters at the White House that he understands companies must regulate themselves. He has opposed slowing the pace of AI development, citing competition with China.

Signatories included Google Chief Executive Sundar Pichai, Anthropic Chief Executive Dario Amodei, Meta Chief Executive Mark Zuckerberg, OpenAI President Greg Brockman, Nvidia Chief Executive Jensen Huang and xAI Chief Executive Elon Musk.

The White House meeting discussed AI safety, industry growth and data center development. Trump ruled out a joint U.S.-China AI venture concept.

Trump also signed an executive order the same day directing the administration to use the term "superintelligence" instead of "artificial intelligence". He argued the term better captures the potential and rapidly advancing capabilities of the technology.