Trump and leadership of major tech companies sign non-binding AI code of conduct
Trump and the heads of Meta, OpenAI, Nvidia, and others signed an AI code of conduct at the White House. It is only morally binding, without legal enforceability, and requires independent audits of model safety.
At the White House, President Trump signed a code of conduct for artificial intelligence together with leadership of major tech companies. According to Politico, the document is only "morally binding" and has no legal weight — what happens if it is violated remains, according to the source, an open question.
According to Reuters, the code was signed by, among others, Meta CEO Mark Zuckerberg, Greg Brockman for OpenAI, Jensen Huang of Nvidia, and Elon Musk. The document requires independent third-party audits to verify that AI models "operate as intended." Oversight of internal security controls aimed at preventing AI models from attacking computer systems should be handled by a separate independent board. According to the quote, Zuckerberg described the code as a starting point, not a final solution.
The move comes amid concerns about uncontrolled AI agents that have made their way onto government websites or launched cyberattacks. Trump also proposed the creation of a ten-member oversight committee, but repeatedly emphasized that he does not want to slow down the development of AI. Critics say that a code without actual legislation is toothless and that similar voluntary commitments have been tried before.
Why it matters
This is a political signal, not an enforceable rule — companies developing AI are committing to independent security audits voluntarily, with no penalties for non-compliance. For companies in the AI field, this may mean growing public and media pressure for transparency in security practices, even though they are not legally compelled to do anything. Whether this will lead to actual legislation is not stated by the sources.
Relevant practical impact
What this means
For a business
Companies developing or deploying AI models may in the future face expectations of independent security audits and oversight of agent behavior, although for now only on a voluntary basis without legal enforceability.
Risks and complianceCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.