Donald Trump gathered the heads of America’s largest artificial intelligence companies at the White House on September 29 and signed what he called a “morally binding” accord on developing the technology safely.

The one-page White House Accord on Super Intelligence was signed by Trump and the chiefs of Google, Anthropic, Meta, OpenAI, Nvidia and XAI. The document was released on the president’s Truth Social platform.

It commits the firms to four layers of voluntary safeguards: internal controls to monitor model behaviour, an internal team to run those controls, an independent external auditor to assess them, and an independent committee of each company’s board of directors to receive reports.

Trump, who has recently dismissed AI dangers as a “hoax”, said the companies should regulate themselves. “There’s a belief that there should be tremendous self-regulation,” he said during the lunch.

Anthropic’s Dario Amodei, among the most outspoken on AI risks, struck a note of caution, saying “the technology has very real risks” and that how to address them is “still under discussion”.