Trump publishes text of “morally binding” AI agreement: external audits and controls for tech giants

President Donald Trump published on Tuesday evening on his Truth Social network the full text of the artificial intelligence agreement he had signed a few hours earlier at the White House with the chief executives of the largest technology companies, as Reuters reports. The one-page document, which Trump described as “morally binding,” envisages four pillars of control that companies must implement in their AI systems.

According to Reuters reporting, the agreement asks companies to cooperate with an external auditor or assessor to carry out independent assessments that AI control and monitoring systems are working properly. It also requires the establishment of an independent board committee, which oversees and receives reports from internal and external auditors and from the teams operating AI controls.

Another point in the document obliges companies to maintain internal controls to monitor the capabilities and “alignment” of models during training and deployment, particularly in areas such as cybersecurity, biosecurity and chemical threats — with the aim that models “do not hack or enter technical systems in unforeseen ways,” as quoted in the agreement text. Companies must also have an internal team ensuring that all controls, monitoring and detection systems are working as designed.

The document bears the signatures of Trump and the heads of Google, Anthropic, Meta, OpenAI, the X platform and NVIDIA, Reuters announced. The text says that “over time, it may make sense to codify these steps into laws or regulations,” adding that the companies believe implementing these controls and audits “is critical to ensuring a safe future for all.” The participating companies will meet regularly to set standards and best practices for the security of their systems.

After the meeting, Trump described the agreement as “it’s almost like a Constitution in a way,” adding that the creation of an oversight committee for the whole sector is being considered. The publication of the text on Truth Social comes at a time when governments around the world are trying to confront the rapid development of artificial intelligence and the warnings of the leading companies themselves about the risks the technology poses to humanity.