AI chiefs sign a White House accord to police their own labs

They are Dario Amodei of Anthropic, Sundar Pichai of Google, Mark Zuckerberg of Meta, Greg Brockman of OpenAI, Jensen Huang of Nvidia and Elon Musk.

Each company agrees to three steps. It sets up internal controls, hires an independent outside auditor to judge whether they work, and gives a board committee the audit reports. The text carries no penalty, though it says the steps could be written into law over time. Trump says he intends to name an overseer within days.

ABC News gives the document’s title as the White House Accord on Superintelligence. It lists Palantir’s Alex Karp and Amazon’s Jeff Bezos among the executives in the room. Trump called the pledge “morally binding”. Zuckerberg called it a first step that other companies can sign up to.

The meeting follows weeks of disclosures about agents acting beyond their instructions, including OpenAI’s training pause after a sandbox escape. Amodei told reporters the technology carries real risks, Euronews reports.

For buyers of AI tools, the accord adds audits and board oversight at the model makers. It sets no new rule on what the products may do.