Trump's AI 'Super Intelligence' Accord Leaves Companies' Promises Voluntary

On Tuesday, top executives from Google, Anthropic, Meta, OpenAI, xAI and Nvidia signed a document called "The White House Accord on Super Intelligence" following a luncheon hosted by President Donald Trump. Trump described the move as an act of "tremendous self-regulation."

However, the agreement falls short of the sweeping AI regulation that industry figures themselves had previously called for. The statement says companies "should implement" several measures: maintaining robust internal controls to monitor their models' capabilities and prevent unintended rogue behavior or hacking; empowering an internal team to handle this oversight and remediation; partnering with an independent external monitor for assessments; and ensuring their board has a committee that receives reports on these efforts.

A voluntary pledge

The commitments are voluntary. Failing to honor them could theoretically be treated as a deceptive practice under the FTC Act, but even then, the typical remedy is simply an order not to make false claims again, not sweeping reform.

This isn't the first such pledge — in early 2025, the UK and South Korea secured similar "Frontier AI Safety Commitments" from the same companies.

FTC investigation looms

Following the luncheon, reports emerged that the FTC is planning a broad investigation into Anthropic, OpenAI, other AI labs, and third-party evaluator METR. An FTC spokesperson confirmed some form of investigation exists but declined to share further details. Douglas Farrar, a former FTC public affairs official, expressed doubt that Trump would allow any serious legal action against companies whose executives he has close ties with.

Google, Anthropic, Meta, OpenAI, xAI, Nvidia and the White House did not respond to requests for comment.