Story
September 30, 2026
Trump’s AI Safety Deal Bets on Tech Giants Policing Themselves
Trump gathered the biggest names in AI behind a voluntary safety accord built around audits and internal controls. Industry leaders called it progress, while critics questioned whether self-policing can match the technology’s risks.
Tuesday began with Trump framing AI dominance as a national imperative. Before hosting technology executives at the White House, he argued the United States could not “stifle growth” and stay ahead of China, while insisting companies must act honestly or face existing law-enforcement powers.1
Hours later, Trump, Speaker Mike Johnson and executives from leading AI firms unveiled the White House Accord on “Super Intelligence,” Trump’s preferred new label for AI. The voluntary pact calls on companies to build internal safeguards, submit systems to outside evaluators and auditors, and have boards independently review audit reports. Trump called the arrangement “morally binding” and said firms would be “policing each other.”2
The administration’s case is that peer scrutiny can protect the public without throttling an industry it sees as strategically vital. David Sacks celebrated Trump’s ability to convene companies developing chips, data centers and frontier models, calling the build-out a new industrial revolution.
3
Participating executives broadly endorsed the direction, if not the idea that it settles the safety debate. Meta’s Mark Zuckerberg described the commitment to “robust internal controls and multiple layers of audits and reviews” as a significant positive step that should bolster confidence in the technology.
4 Google CEO Sundar Pichai said the group had signed the accord after what he called an important conversation with Trump administration officials and industry leaders.
5
Yet the deal’s limits were apparent from the outset. Its full enforcement structure and scope had not been released, and AI-industry sources told Axios that self-policing may not solve the safety problem. Anthropic chief Dario Amodei, who signed the agreement, said the technology carries “very real risks” and that the mechanism for addressing them remained under discussion.1
That leaves the accord as both a display of unusual industry unity and a test of whether voluntary promises can keep pace with systems becoming more powerful—and more consequential.