Trump posts White House Accord on Super Intelligence
The voluntary pact asks AI companies to run internal audits on model capability and safety, with no enforcement mechanism attached
Published
President Trump posted the "White House Accord on Super Intelligence" on Tuesday, a joint commitment among frontier AI companies on what the accord calls their shared responsibilities. It calls for participating companies to meet regularly to set standards and best practices.
The accord asks each company to put in place four layers of controls and audits. These include internal monitoring of a model's capabilities and its alignment with safety goals in areas such as cybersecurity, biosecurity and chemical threats, along with steps meant to stop models from hacking or accessing technical systems on their own.
The text notes that it may make sense over time to write these steps into law or regulation, but says implementing the controls now does not depend on that happening. That framing leaves the legal status of the commitments open rather than fixed.
The accord is voluntary and carries no enforcement mechanism. It follows a run of White House engagement with AI companies that has leaned toward access rather than restriction, including a meeting between Trump and Anthropic's Dario Amodei on Monday and an AI-focused lunch with Elon Musk and Jeff Bezos on Tuesday morning. Because compliance is self-directed and the audits described are internal rather than externally mandated, the near-term operating picture for signatories such as OpenAI and Anthropic is unchanged. The accord reads as a formalizing of safety messaging already in place rather than a new constraint on model development or deployment timelines, and the net effect on AI-exposed equities stays neutral.