Trump, tech CEOs sign AI code of conduct at White House

The signing at the White House brings together leaders from Meta, OpenAI, Nvidia, and Elon Musk, but the document they signed is not legislation. According to Politico, it is only "morally binding," and what happens if someone violates it remains an open question. That gap is why critics call a code of conduct without actual legislation toothless, noting similar voluntary commitments have been tried before.

The code does include concrete mechanisms: independent third-party auditors are supposed to verify that AI models are "operating as intended," and a separate independent board would oversee internal safety checks designed to prevent AI models from hacking into systems. Zuckerberg framed the code as a starting point, not a final solution, which suggests signatories see it as an early step rather than a finished framework.

The backdrop is growing concern about uncontrolled AI agents that have accessed government websites or launched cyberattacks. Trump has proposed a ten-member oversight committee but stressed repeatedly that he doesn't want to slow AI growth, and he used an executive order to rename AI "Super Intelligence." Whether the voluntary audits and board carry weight may hinge on whether signatories treat the code as a real constraint or a symbolic gesture.