Here’s what AI leaders are saying about Trump’s new safety plan

Jess Weatherbed reports that after a meal with tech leaders, President Trump framed AI safety fears he once called a hoax as outdated now that he’s talking about “Super Intelligence,” and pointed to U.S. companies “dominating China” instead of new federal guardrails. The signees of the Joint Commitment on Frontier Responsibilities are Sundar Pichai, Dario Amodei, Mark Zuckerberg, Greg Brockman, Elon Musk, and Jensen Huang.

The Verge describes the pledge as morally binding self-regulation with common-sense guidelines and seemingly no ramifications for violations. Executives offered on-camera praise for advancing AI, jobs, and public opinion. The piece is a readout of the political theater around the same voluntary package covered elsewhere as an accord.

We’re watching the quotes, not the applause lines.

When the President says the smartest people in the world are watching over each other, that’s a governance theory — peer reputation as regulator. Sometimes that works inside a profession. It fails when competitive pressure and product calendars punish the cautious lab.

I’m listening for what these leaders refuse to promise: timelines, penalties, independent publishable audits. Praise for jobs and “negative public opinion” is politics. Useful AI still needs someone who can say no to a launch.

We’ll cover their statements fairly. We’ll also keep a parallel ledger of incidents and gated models that show why self-policing alone is a thin shield.

The Verge, September 30, 2026 (Jess Weatherbed).