Trump Meets AI CEOs at White House, Pushes 'Super Intelligence'
On September 29, 2026, President Trump hosted a White House luncheon on what he calls "super intelligence". Leaders of Google, Anthropic, Meta, OpenAI, xAI and NVIDIA signed the White House Accord on Super Intelligence, a voluntary pledge to police their own frontier AI. Trump called it "morally binding". Updated October 30, 2026: here is what the accord actually says, who signed, what has happened since, and what it does not do.
⚡ Quick facts
- Event: a White House luncheon on super intelligence, September 29, 2026
- Signatories: Trump plus Sundar Pichai, Dario Amodei, Mark Zuckerberg, Greg Brockman, Elon Musk and Jensen Huang
- What it requires: four layers: internal monitoring, an internal verification team, an independent external auditor and board oversight
- Legal status: voluntary; the text says it “may make sense” to codify the steps into law over time
- Follow-up: a White House AI task force launched in early October to put the accord into practice, per CFR
What the accord says
The full text, archived by the American Presidency Project, says each company is responsible for developing its own technology safely. Firms that train or deploy frontier models commit to four layers of controls: monitoring of capabilities and alignment during training and deployment (covering cyber, biological and chemical risks); an internal team that checks the controls work; an independent external auditor or evaluator; and an independent board committee that receives both teams' reports. The companies also agree to meet regularly to set standards.
What it does not do
It creates no regulator, no reporting duty to government and no penalty. The companies say they will follow the controls whether or not the law requires it, and the text only says that "over time, it may make sense to codify these steps into laws or regulations." Per CNBC, Trump called it "morally" binding, and House Speaker Mike Johnson described it as voluntary. It also does not rename AI: "super intelligence" is Trump's preferred term, not an official change.
Who was there, and the tension behind it
The White House hosted the event in the East Room. Anthropic's Dario Amodei, who has argued for a slower, more careful pace at the frontier, signed alongside companies that favour faster development. He had a private dinner with Trump the night before, and afterwards said AI risk rules were "still under discussion". Forbes published the accord text the next day, noting how short it is on specifics such as who picks the auditors.
What happened next
In early October, the White House launched an AI task force to put the accord into practice. Per the Council on Foreign Relations, Anthropic, Google, Meta, NVIDIA, OpenAI and SpaceXAI committed to it. The task force is meant to identify AI security threats, especially from foreign adversaries, and help shape safety standards while leaving industry to lead.
Our take: how to judge a voluntary AI pledge
- Who audits? The accord requires an independent auditor but does not say who qualifies or who chooses them.
- What gets published? Nothing in the text requires public reports. Without disclosure, outsiders cannot check compliance.
- What happens on a breach? There is no penalty. A pledge without consequences depends on reputation alone.
- Does it pre-empt stronger rules? Watch whether it is used to argue against state or federal laws, such as the proposals in our Superintelligence Ban Act explainer.
- Does it reach beyond the US? These companies serve users worldwide, so their internal controls apply globally, but no other government is a party to it.
For what binding oversight can look like, see our explainer on the FTC's investigation into rogue AI agents.
Sources
- American Presidency Project: White House Accord on Super Intelligence (full text)
- The White House: luncheon on Super Intelligence, September 29, 2026
- CNBC: Trump says tech leaders signed 'morally binding' AI agreement
- Forbes: what the White House AI accord says
- Council on Foreign Relations: how to make the new AI task force a success
Frequently asked questions
What is the White House Accord on Super Intelligence?
A voluntary pledge signed on September 29, 2026 by President Trump and the heads of Google, Anthropic, Meta, OpenAI, xAI and NVIDIA to apply four layers of safety controls to frontier AI.
Who signed the accord?
Donald Trump, Sundar Pichai (Google), Dario Amodei (Anthropic), Mark Zuckerberg (Meta), Greg Brockman (OpenAI), Elon Musk (xAI) and Jensen Huang (NVIDIA).
What are the four layers of controls?
Internal monitoring of model capabilities and alignment, an internal team that verifies those controls, an independent external auditor, and an independent board committee that oversees both.
Is the accord legally binding?
No. It is voluntary and has no penalties. Trump called it morally binding, and the text says the steps may be codified into law over time.
Did Trump rename artificial intelligence?
No. Trump said he prefers the term super intelligence, but there has been no official change in terminology.
What happened after the accord was signed?
In early October 2026 the White House launched an AI task force to put the accord into practice, with Anthropic, Google, Meta, NVIDIA, OpenAI and SpaceXAI committing to it.