Details of Trump-Backed AI Safety Deal Show Self-Policing Rules, No Penalties

The accord, officially titled the Joint Commitment On Frontier Responsibilities, was signed by Google's Sundar Pichai, Anthropic's Dario Amodei, Meta's Mark Zuckerberg, OpenAI's Greg Brockman, XAI's Elon Musk, and Nvidia's Jensen Huang. Trump also added his own signature, with his title listed in the document as "President of the Unites States," the misspelling included as written. The deal was announced a day earlier.

The document opens by saying that "in order to build a positive future for the American people and the world, we believe every company is responsible for developing its own technology safely and in a way that builds trust with customers and the public." It adds that the participating companies "will meet regularly to establish standards and best practices to improve the safety of their systems."

The agreement outlines four rules for each company, aimed at ensuring frontier AI models behave as intended and that any problems are quickly identified and resolved. The first asks companies to implement robust internal controls to monitor the capabilities and alignment of their models during training and deployment in areas including cybersecurity, biosecurity, and chemical threats, and to ensure that models do not hack or access technical systems in unintended ways. The second calls for empowering an internal team to ensure all controls, monitoring, and detection are operating as intended and that any issues are remediated. The third requires partnering with an independent external auditor or evaluator to carry out independent assessments of whether those controls, monitoring, and detection are working as intended. The fourth asks companies to designate an independent committee of the board of directors to oversee and receive reports from the teams operating the controls and the internal and external auditors and evaluators, as well as to ensure any issues identified are remediated.

The agreement says these steps will give the companies, their users, and the general public, which has mounting concerns around AI safety, "confidence that the technology is operating as intended." A later passage also says that "over time, it may make sense to codify these steps into laws or regulations."

The Verge reported that the rules are mostly vague, common-sense safety procedures for any technology company and do not mention any penalties for violating the agreement, making the commitment little more than a pinky promise from AI leaders to do the bare minimum. The publication also noted that Google, OpenAI, and Anthropic have all violated that first rule over the last few months when their AIs went rogue and hacked other companies and even government websites, according to its report.

Trump introduced the deal at the same time as ordering the US government to refer to AI as "Super Intelligence" going forward, reasoning that the word "artificial" sounds bad and that "super is the best word of all," The Verge reported. Alongside the technology executives, Trump signed the AI safety agreement himself.

Editor's Summary

President Trump's AI safety agreement, signed by leaders from Google, Anthropic, Meta, OpenAI, XAI, and Nvidia, sets four self-regulatory procedures for frontier models, including internal controls, internal and external audits, and board oversight. The document contains no penalties for violations, and The Verge described it as a vague, common-sense pledge. Trump simultaneously directed the US government to call AI "Super Intelligence."