Trump's AI Safety Deal: Tech CEOs Sign 'Morally Binding' Pledge to Self-Regulate

President Trump's administration has released the full text of a voluntary AI safety accord signed by leading technology executives, titled the Joint Commitment on Frontier Responsibilities. The document, posted online by tech founder and presidential adviser David Sacks, commits companies to self-regulate their most advanced artificial intelligence systems. The deal is described as "morally binding" and carries no legal penalties for noncompliance, according to the text.

The commitment outlines a set of principles for developing so-called frontier models, the largest and most capable AI systems. Signatories agree to conduct internal risk assessments, share certain safety findings with peers, and maintain human oversight of high-stakes deployments. The accord does not create an independent regulator or require third-party audits, relying instead on companies to hold themselves accountable. Specific deadlines or enforcement mechanisms are absent from the published document.

Sacks, who serves as the White House's AI and crypto czar, shared the details on social media, framing the agreement as a lighter-touch alternative to previous regulatory efforts. The signatories include executives from several top AI labs, though the full list was not immediately clear from the initial release. The move follows Trump's repeal of a Biden-era executive order that had required companies to report safety test results to the federal government.

Industry reaction has been mixed. Some safety researchers argue that without government oversight, the pledge amounts to little more than public relations. Others in the tech sector welcome the flexibility, saying rigid rules could slow innovation and cede ground to competitors abroad, particularly China. The accord's moral framing echoes earlier voluntary AI principles, such as the 2023 Bletchley Declaration, which also lacked binding enforcement.

What the deal changes in practice remains uncertain. Companies that sign on face no fines or sanctions if they fail to meet the commitments, and there is no neutral body to verify their claims. The White House has not said whether it will track compliance or publish progress reports. For now, the burden of proof falls on the companies themselves to demonstrate that self-regulation is more than a slogan.

The coming months will test whether the pledge influences actual development decisions. Watch for whether major labs disclose safety incidents publicly, and whether Congress moves to codify any of these principles into law. The open question is whether a morally binding handshake can hold when competitive pressures and billions of dollars are on the line.