Donald Trump and Heads of US Tech Giants Sign Agreement on AI Oversight

US President Donald Trump met with the heads of leading American technology companies—Google, Anthropic, Meta, OpenAI, xAI and Nvidia. Following the meeting, an agreement on joint commitments regarding artificial intelligence oversight was signed. The document, titled the “White House Superintelligence Agreement,” was published on the US president’s Truth Social account.

Trump said the agreement has “moral force” and is not subject to judicial enforcement. However, the document states that “over time, it may become necessary to enshrine these measures in laws or regulations.”

The document states that each company is responsible for developing technology safely. To this end, developers “must have robust internal processes and controls ensuring that the behavior of their technologies is consistent with their intended purpose and that any problems are promptly identified and resolved.”

The agreement provides for four levels of oversight and auditing for companies developing advanced AI models:

• Establishing robust internal controls to monitor the capabilities and compliance of their models during training and deployment in areas such as cybersecurity and biological and chemical threats, and to ensure that the models do not hack technical systems;
• Maintaining an internal team to ensure that all control, monitoring and detection mechanisms work as intended and that any problems are promptly resolved;
• Engaging an independent external auditor or evaluator to independently assess whether the control, monitoring and detection mechanisms are working properly;
• Appointing an independent board committee to oversee the teams managing the control mechanisms, receive reports from internal and external auditors and evaluators, and ensure that identified problems are resolved.

The document also states that the companies will meet regularly to establish standards and best practices for improving the safety of their systems.