Trump and AI leaders sign voluntary safety pledge for frontier models

The deal puts major AI labs in charge of their own safeguards, with outside audits and board oversight instead of immediate regulation.

President Trump and several leading AI executives have signed a voluntary frontier-AI safety pledge that asks companies to police their own systems, a move that matters because it could shape how the biggest labs handle security without new regulation. The agreement, called the Joint Commitment on Frontier Responsibilities, requires firms to add internal controls, use outside auditors, and create board-level oversight for risks such as cybersecurity, biosecurity, chemical threats, and unintended model behavior. Signers include leaders from Google, Anthropic, Meta, OpenAI, xAI, Nvidia, and Trump himself, and the document says the companies will meet regularly to set shared standards and best practices. The administration is presenting the deal as a “morally binding” alternative to heavier-handed rules, even though the text itself leaves enforcement vague and suggests the steps could be written into law later.