Top AI firms sign 'morally binding' pledge at White House to police themselves
Six tech leaders promised internal controls and outside audits of their most advanced models. The agreement is voluntary and has no enforcement.
The heads of six leading AI and chip companies signed a voluntary commitment at the White House on Tuesday to audit and control their most powerful models.
The leaders of six of the most powerful AI and chip companies signed a voluntary pledge at the White House on Tuesday to police the development of their most advanced models, as President Donald Trump ruled out formal government guardrails.
The signatories were Anthropic's Dario Amodei, OpenAI's Greg Brockman, Google's Sundar Pichai, Meta's Mark Zuckerberg, xAI's Elon Musk and Nvidia's Jensen Huang, according to Al Jazeera and Fortune.
The agreement is the administration's main answer to growing alarm about AI systems acting in ways their makers did not intend. But it is non-binding and has no enforcement mechanism. Trump called it "morally binding."
"I think I'm seeing tremendous self-policing. And they understand that they have to self-police," Trump said.
The companies committed to four things: "robust" internal controls, independent outside auditors, board-level committees to review the auditors' findings, and regular meetings to develop shared safety standards, according to Fortune and Euronews.
The controls are meant to watch for cybersecurity, biological and chemical threats, and to make sure systems cannot "hack or access technical systems in unintended ways," Fortune reported.
The meeting came after a string of incidents involving AI agents behaving unexpectedly. OpenAI halted a model rollout over safety concerns the day before, AP reported, and Al Jazeera and Euronews said the company disclosed that one of its agents had broken into an Australian government website during testing.
"The technology has very real risks," Amodei said.
Trump made clear he would not go further. "I will never stifle the growth of a technology that will be bigger than the industrial revolution," he said. The text itself says it "may make sense to codify these steps into laws and regulations" over time.
Critics said a voluntary pledge falls short. "What other trillion-dollar industry marks its own homework?" Toby Walsh of the AI Institute at the University of New South Wales told Al Jazeera.
Reports differ on what the document is called, and its full official text had not been published on the White House science office's website as of this week.
Other leaders attended the lunch but are not listed as signatories, including House Speaker Mike Johnson, Amazon's Jeff Bezos and Palantir's Alex Karp, ABC News reported. It is not clear whether companies that did not sign will join.
Trump said a committee of about 10 people would "watch over the whole enterprise," and that he would name an overseer in the coming days. Fortune reported that Google, OpenAI and Anthropic plan to launch an industry-run standards body by early 2027.