Late on September 29, President Trump released the text of the AI accord that leaders of six AI companies signed at the White House. It asks each company to watch its models for dangers such as cyber and biological threats, to hire an independent auditor, and to have an independent board committee oversee the checks. The companies will “meet regularly” to set standards.
New to this? Read it in simple words
- President Trump shared the AI promise that big tech bosses signed at the White House.
- The companies promise to check their AI and to let outside experts check it too.
- There are no deadlines or punishments if a company does not keep the promise.
- The text says these steps might become law one day.
- Auditor
- An outside expert who checks whether rules are really followed.
- Board committee
- A small group of a company’s directors who oversee one area.
- Codify
- To write a rule into law.
The four layers
The document is titled “White House Accord on Super Intelligence Joint Commitment on Frontier Responsibilities”. It says every company that trains and deploys frontier models should put in place “four layers of controls and audits”.
The first is internal controls that watch a model’s capabilities and alignment during training and use, in areas such as cybersecurity, biosecurity and chemical threats. They should also ensure that models “do not hack or access technical systems in unintended ways”.
The second is an internal team that makes sure the controls work and that problems are fixed. The third is an independent outside auditor or evaluator. The fourth is an independent committee of the company’s board, which receives the reports.
The four layers as the released accord lists them.
Who signed, and what is missing
Six executives signed: Google’s Sundar Pichai, Meta’s Mark Zuckerberg, Anthropic’s Dario Amodei, Elon Musk, Nvidia’s Jensen Huang and OpenAI’s president, Greg Brockman. Trump said he had signed it as president.
Microsoft’s Satya Nadella and AMD’s Lisa Su were at the White House that day, CBS News reports, but they are not among the signers.
The text sets no deadlines and no penalties, and asks for no reports to the government. The companies “will meet regularly to establish standards and best practices to improve the safety of their systems”, it says.
Law later?
“Over time, it may make sense to codify these steps into laws or regulations,” the accord says. The signers add that they are committed to the controls whether or not the law requires them.
Senator Mark Warner, a Democrat, urged Congress to pass mandatory testing and incident-reporting rules for the most advanced models. “The president’s response? To rename it and tell the companies developing it to regulate themselves,” he said, according to CBS News.
Sources
Every fact in this story comes from the sources below. Open them to check our work.
- 1Research · September 29, 2026READ IN FULL: Trump and tech leaders’ White House Accord on Super Intelligence Washington Examiner
- 2
- 3Research · September 29, 2026Trump and major AI executives sign “morally binding” voluntary controls: “It’s almost like a constitution” CBS News
Trump posted the text as an image on Truth Social, which we could not open without an account. We read the full text as printed by the Washington Examiner and checked it against the passages Reuters and CBS News quote from the document.