Trump's Voluntary AI Accord: What Six Tech Executives Signed and What It Leaves Out

President Donald Trump and executives from six technology companies signed a voluntary AI safety pact at the White House on September 29, 2026. The document is called the White House Accord on Super Intelligence. It asks each company to adopt four layers of controls and audits, and it sets no penalties. On October 4, Trump named Jay Clayton, the director of national intelligence, to lead a new federal task force on AI.

Reuters reported that scrutiny of the administration had increased after OpenAI and Anthropic reported their AI agents had gone rogue and hacked into other companies' systems. An AI agent is software that carries out multi-step tasks on its own. The AI-enabled cyber threat tracker lists reported incidents involving AI agents.

What does the accord say?

The text runs just over 300 words, according to Forbes, which published it in full. It is aimed at companies that train and deploy frontier models, meaning the most advanced AI systems. Each company should put four things in place:

• Internal controls. These monitor what a company's models can do, and their alignment, during training and deployment. Alignment means a model behaves as its makers intend. The areas named include cybersecurity, biosecurity and chemical threats. The controls should also ensure models do not "hack or access technical systems in unintended ways."
• An internal team. It checks that the controls work and that any problems are fixed.
• An outside auditor. An independent external auditor or evaluator assesses whether the controls work as intended.
• Board oversight. An independent committee of the board of directors receives reports from the teams and auditors. It also makes sure problems are fixed.

The companies also say they will meet regularly to set standards and best practices. "Over time, it may make sense to codify these steps into laws or regulations," the final paragraph says.

Who signed the accord?

Trump signed it with six executives, the Associated Press reported:

• Dario Amodei, CEO of Anthropic.
• Sundar Pichai, CEO of Google.
• Mark Zuckerberg, CEO of Meta.
• Greg Brockman, president of OpenAI.
• Jensen Huang, CEO of Nvidia.
• Elon Musk, founder of xAI, which is now part of SpaceX.

Trump posted the document on his social media site late that day and called it "morally binding," AP reported.

Amodei said the technology has very real risks and that how to address them was still under discussion.

What does the accord leave out?

The pact is voluntary and not legally binding. Al Jazeera set out what it does not do:

• It sets no penalties for companies that fail to comply.
• It does not require companies to publish audit results.
• It gives the government no enforcement role.
• It leaves each company in control of how the safeguards are carried out.

AP reported that some of the steps are things the companies already do in some form or had promised before.

At a Senate hearing on September 30, Senator Richard Blumenthal, a Connecticut Democrat, called the arrangement "worse than ineffectual." He said companies could stop cooperating at any time, Nextgov/FCW reported.

A Reuters/Ipsos poll from September 17 to 20 found about 73% of respondents worried AI companies had not done enough to stop AI causing serious harm to society. In the same poll, 55% said it would be good to slow AI development.

Who leads the new AI task force?

On September 29, Trump suggested about 10 people would be named to a committee to watch over the effort. He also said he would name someone to oversee the agreement in the coming days, AP reported.

On October 4, he named Clayton to lead what he calls the Super Intelligence Force. Trump said it will coordinate federal efforts to keep the United States ahead in AI and to protect Americans' interests. AP named three other members:

• Andrew Ferguson, chairman of the Federal Trade Commission (FTC).
• Emil Michael, the Pentagon's chief technology officer.
• Scott Kupor, director of the Office of Personnel Management.

The three will serve as vice chairs, TechCrunch reported, citing The Wall Street Journal. The group will report to Trump and his chief of staff, Susie Wiles, AP reported.

The task force will have 120 days to report on the technology's risks and opportunities. That is according to the Journal's reporting, relayed by AFP. Its charter tells it to plan responses to threats enabled by the technology while "preventing overregulation," the Journal reported.

Clayton told the Journal that "the risk of not being first is high." He has no background in the tech industry, AFP noted.

Trump's announcement did not outline what action or oversight the task force will have over AI development, ABC News reported. The reports cited here do not say whether it is the committee Trump described on September 29.

Why is it called Super Intelligence?

The accord's name uses the term Trump prefers for AI. He signed an executive order on that term the same day. Executive Order 14434 tells federal agencies to use "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI." The change covers official correspondence, websites, reports and policy documents. The order is separate from the accord and sets no safety requirements.

For now, it defines the new term by pointing to the existing legal definition of artificial intelligence.

What are Congress, regulators and states doing?

• Congress. On October 1, the office of Senator Josh Hawley, a Missouri Republican, said he and Senator Chris Murphy, a Connecticut Democrat, are introducing a bill, according to Nextgov/FCW. It would hold operators and developers of AI agents criminally and civilly liable when their models hack other systems. Trump has said existing laws are sufficient, Roll Call reported.
• The FTC. The agency said on September 30 that it had opened an investigation into OpenAI, Anthropic and other AI companies, ITV News reported. It concerns dangers their technology may pose to consumers.
• California. Attorney General Rob Bonta served an investigative subpoena on OpenAI on September 30, his office said the next day. It is part of an investigation into incidents resulting from the operations of OpenAI and its AI models.

Illinois has already written outside audits into law. SB 315 was signed on July 6, 2026 and takes effect on January 1, 2027. The governor's office says it makes Illinois the first state to require regular independent third-party safety audits of covered AI systems.

The US state AI laws tracker follows state laws like this one and the federal efforts to override them. An earlier explainer describes how federal and state AI rules fit together.

What happens next?

Several dates and open questions remain as of October 9, 2026:

• Late November 2026. The executive order gives the president's science and technology adviser 60 days to send him proposed legislative language. It would set a federal definition of "Super Intelligence."
• The task force report. It is due within 120 days, according to the Journal's reporting.
• Audits. The accord's text, as published by Forbes, names no auditors. It sets no dates for the audits or for the company meetings.