Trump and six tech chiefs sign a voluntary accord on AI safety controls

After a lunch at the White House, President Donald Trump and six technology chief executives signed the White House Accord on Super Intelligence, a one-page set of voluntary commitments on how companies that train frontier models should police them. The signatories were Sundar Pichai of Google, Dario Amodei of Anthropic, Mark Zuckerberg of Meta, Greg Brockman of OpenAI, Elon Musk of xAI and Jensen Huang of Nvidia. OpenAI’s chief executive, Sam Altman, was at the company’s DevDay conference in San Francisco. Microsoft’s Satya Nadella and Amazon founder Jeff Bezos attended the lunch, CNBC reported, but are not among the signatories. The accord was signed on the same day as Trump’s order renaming AI “super intelligence” in federal usage.

The document commits each company to “four layers of controls and audits”:

(1) Implement robust internal controls to monitor the capabilities and alignment of its models during training and deployment around areas like cybersecurity, biosecurity, and chemical threats, and to ensure that its models do not hack or access technical systems in unintended ways.

— White House Accord on Super Intelligence

The other three layers are an internal team to check the controls are working, an independent external auditor or evaluator, and an independent committee of each company’s board to receive the auditors’ reports. The companies also agreed to “meet regularly to establish standards and best practices”. The text says it “may make sense to codify these steps into laws or regulations” over time.

Asked whether the accord was binding, Trump said: “I think it’s morally binding.” He called it “almost like a constitution” and said he was “seeing tremendous self-policing”, CBS News reported. House Speaker Mike Johnson described it as a statement of principles that were “voluntary on behalf of the industry”. Trump said the group had discussed a ten-person committee to “watch over the enterprise” and that he would name an AI czar within days. Amodei told reporters that rules addressing the risks he had raised were “still under discussion”, according to CNBC.

The accord followed a summer of agent incidents at OpenAI and Anthropic and weeks of calls from Amodei and Altman to slow frontier development, which Trump had called a “hoax”. Like the 2023 Biden-era commitments, it contains no enforcement mechanism; unlike them, it asks for oversight by a committee of each company’s board. The commentator Zvi Mowshowitz argued that the clause on meeting to set standards came “close to an antitrust waiver” in practice, at a time when a subscriber lawsuit was calling the labs’ pacing pledge a cartel.