A two-letter slip is overshadowing the substance of President Donald Trump’s new artificial intelligence pact: beneath the president’s signature on the White House Accord on Super Intelligence, signed with six technology executives at the White House on Sept. 29, 2026, the document reads “Unites States.”
The agreement was posted to the White House’s social media account Sept. 30 and spread rapidly online with the misspelling intact. How the error got into the signature block is not clear, nor is it clear whether the version that appears in the Federal Register will look the same.
Social media users and politicians, among them California Gov. Gavin Newsom, seized on the flub within hours.
What the Accord Asks of AI Companies
Trump has described the pact, which he calls “morally binding,” as a landmark moment of cooperation between Washington and Silicon Valley. “It’s almost like a constitution in a way,” Trump said of the signing.
The executive signatories are Google CEO Sundar Pichai, Anthropic CEO Dario Amodei, Meta CEO Mark Zuckerberg, Greg Brockman of OpenAI, Nvidia CEO Jensen Huang and SpaceX CEO Elon Musk. The document, also referred to as the Joint Commitment on Frontier Responsibilities, was announced after a White House luncheon with top tech executives, and Trump posted it to his Truth Social account on Tuesday, Sept. 29.
Its core ask is that the companies maintain strong internal processes and controls so their systems behave the way their builders intend. It also tells signatories to work with an independent external auditor and to create a board-level oversight committee — a requirement the text renders as designating a committee “to oversee to receive reports, oversee and receive reports from the terms operating the controls.”
Four Layers, No Regulators
The agreement lays out four layers of controls and audits. The first is internal safety monitoring during model training, a practice that already exists at the major labs and that has not prevented testing failures to date. The second empowers an internal team to confirm that the first layer is working. The third gives an outside auditor the ability to independently assess a company’s safety controls. The fourth asks each firm to designate an independent board to review reports on those controls.
What the framework does not include is government. None of the listed controls involve federal regulators, and the text carries no commitment to publicly release the findings of independent evaluations. Companies appear free to choose their own evaluators, appoint their own oversight boards and decide for themselves whether to publish results. The pact carries no enforcement mechanism and no legal weight; it is a voluntary commitment, and Trump has pitched it as a self-policing arrangement.
The industry’s safety record is under fresh public scrutiny following a series of incidents in which major AI companies’ products hacked into outside organizations during safety testing. Lawmakers, researchers and even some AI firms have pushed for more government oversight, but Trump has broadly rejected calls to slow down, citing competition with China and the risk to economic growth. The tech sector, which has largely aligned itself with his administration, has lobbied against strict regulatory oversight on the grounds that it would hobble the economy.
Rebranding AI as ‘Super Intelligence’
The accord was one of two moves Trump made Sept. 29. He also signed an executive order replacing the term “Artificial Intelligence” with “Super Intelligence” throughout the executive branch, directing every department and agency to use “Super Intelligence” and “SI” in official correspondence, public communications, policy documents and non-statutory documents — and to stop acknowledging the terms “Artificial Intelligence” and “AI” altogether. The order tasks the Assistant to the President for Science and Technology (APST) with proposing a federal definition of the new terms and identifying any further executive action needed to put it in place.
A White House fact sheet titled “President Donald J. Trump Inaugurates The Era of Super Intelligence,” dated Sept. 29, argues the new name better captures what the technology can do and says the United States has led the field for more than 70 years. It also credits Trump with eliminating what it calls the Biden administration’s overreaching and harmful SI policies immediately upon his return to office, and points to more than $5 billion announced for the Genesis Mission, a national push to accelerate breakthroughs in science, medicine, energy and manufacturing, plus the more than 90 federal actions identified in America’s AI Action Plan, released in July 2025.
Earlier on Sept. 29, Trump previewed the rebrand at the Andrew W. Mellon Auditorium in Washington, D.C., where he unveiled America.gov, a website billed as an AI-powered one-stop shop for federal services and information. He had teased the “super intelligence” label at the United Nations General Assembly. For now, the most-shared line from the whole rollout is not a policy provision — it is the two letters missing from the president’s own country.