Trump Announces Voluntary AI Agreement After White House Meeting
President Donald Trump said Tuesday that he and executives from major artificial intelligence companies signed a voluntary agreement to review the technology’s development. Speaking after a White House meeting, Trump described the approach as industry “self-policing” and said the agreement includes internal and external reviews. The account did not specify how the reviews would be conducted or enforced.
At the same time, Trump said the federal government would not limit AI development or endorse additional safety restrictions. He framed the technology as a major economic opportunity and described its development as a competition with China. Earlier Tuesday, his administration introduced America.gov, an AI chatbot designed to help people navigate government services. Secretary of State Marco Rubio said the service could eventually support fully online first-time passport applications.
The meeting brought together executives and founders associated with Anthropic, OpenAI, Amazon, xAI, Nvidia, Google, Meta and Microsoft. Their companies are investing heavily in AI infrastructure, including data centres that have faced opposition in some communities. Trump said executives had discussed ways to address those concerns, including support for schools and measures to reduce energy costs.
Views on the pace of development differed among participants. Anthropic CEO Dario Amodei and other industry leaders have called for more time to assess safety risks. OpenAI CEO Sam Altman said his company had paused a model rollout over safety concerns. Other executives emphasized companies’ responsibility for the safety of their own products.
Canadian-British AI researcher Geoffrey Hinton, speaking to CBC News before the agreement was announced, criticized Trump’s understanding of AI and called for federal oversight through an agency comparable to the Food and Drug Administration. His comments reflected a broader disagreement over whether voluntary commitments are sufficient to manage the technology’s risks as it advances. No timeline for implementing the voluntary agreement was announced Tuesday.