Trump’s AI Accord Promises Safety While Leaving the Industry to Police Itself

Trump’s AI Accord Promises Safety While Leaving the Industry to Police Itself
The White House has recast AI oversight as a test of trust: can the companies building increasingly powerful systems safely police themselves, or does voluntary restraint simply defer the hard choices to Washington?

President Donald Trump and major technology executives signed a voluntary accord promising internal controls, outside audits and board oversight for frontier AI systems. Trump called the document “morally binding” and compared it to “a constitution,” while House Speaker Mike Johnson described it more cautiously as a statement of principles and “commitments that are voluntary on behalf of the industry.”

The agreement’s supporters frame that flexibility as a strength. It requires companies to monitor models for cybersecurity, biosecurity and other risks, appoint internal oversight teams, commission independent evaluations and report to board committees. The accord also says that, “over time, it may make sense to codify these steps into laws or regulations.” That language leaves open a path toward regulation without slowing the industry immediately—an approach aligned with Trump’s argument that a development pause would undermine American competitiveness.

Yet the central criticism is that the safeguards remain unenforceable. Senator Mark Warner said companies were warning that “the technology is advancing faster than our safeguards,” while Trump’s response was “to rename it and tell the companies developing it to regulate themselves.” The administration simultaneously ordered agencies to use “super intelligence” rather than artificial intelligence, a rhetorical change that does not alter the underlying risks.

Even some participants acknowledged the danger. Anthropic CEO Dario Amodei said AI presented “very real risks” and that the mechanism for addressing them was “still under discussion.” That convergence is significant: executives and officials agree safety matters, but not on who should enforce it. The accord offers process and optimism; its critics see a promise without penalties, leaving the companies responsible for marking their own homework.