Former President Donald Trump and several major tech executives signed a voluntary, morally binding pact on AI safety. The accord grew out of conversations between Meta CEO Mark Zuckerberg and House Speaker Mike Johnson. It emphasizes industry self-policing rather than government regulation.
Zuckerberg was reportedly instrumental in drafting and circulating the proposal, while Nvidia CEO Jensen Huang helped rally support from other industry leaders. The agreement calls for companies to use four layers of controls, including internal monitoring, independent external audits and board-level oversight, to help ensure AI systems are safe and operate as intended.
Other signatories included leaders from Google, OpenAI, Anthropic and xAI.