President Donald Trump announced Tuesday that leading artificial intelligence firms signed a voluntary safety accord, a move designed to calm public anxiety about the rapidly advancing technology without imposing government mandates.
The agreement emerged from a White House meeting where Trump and House Speaker Mike Johnson gathered chief executives from companies pouring hundreds of billions into AI infrastructure. That spending spree has ignited bipartisan backlash in communities nationwide over sprawling data centre projects.
Trump characterized the arrangement as evidence of “tremendous self-policing” during an impromptu West Wing press appearance. The accord, posted on Truth Social, outlines four voluntary commitments: strong internal controls, independent external audits, board-level oversight committees, and a pathway toward future regulation.
“Over time, it may create sense to codify these steps into laws and regulations,” the document states.
Signatories included Anthropic CEO Dario Amodei, Google’s Sundar Pichai, Meta’s Mark Zuckerberg, OpenAI President Greg Brockman, Nvidia’s Jensen Huang, and Elon Musk representing xAI. Amodei, among the industry’s most vocal advocates for slowing AI development, acknowledged the technology carries “very real risks” even as mechanisms for addressing them remain unsettled.
Trump promised to name roughly 10 people to a committee that would monitor the enterprise. He called the deal “morally binding” and said a designated overseer would emerge in coming days.
Earlier Tuesday, the president unveiled America.gov, an AI chatbot intended to help citizens navigate government services. Secretary of State Marco Rubio noted the tool could soon enable fully online first-time passport applications, eliminating in-person visits.
Meanwhile, OpenAI halted rollout of a new model Monday over safety concerns. CEO Sam Altman told CNBC his company now paces its progress deliberately. “We have to do these in a way where we can create very confident safety cases,” he said.
Trump framed AI development as a winner-take-all contest with China, rejecting cooperation despite a recent bilateral agreement on incident response channels. The coming months will reveal whether voluntary commitments satisfy critics or fuel demands for binding oversight.
