Trump Just Renamed AI “Super Intelligence” — Here’s What Actually Changed

Last updated: October 2, 2026

Trump Just Renamed AI “Super Intelligence” — Here’s What Actually Changed

On Tuesday, 29 September 2026, Donald Trump did two things at once: he signed an executive order instructing the entire US federal government to stop using the words “artificial intelligence” and start saying “Super Intelligence” instead — and he got the bosses of Anthropic, OpenAI, Google, Meta, xAI and Nvidia to sign a voluntary AI safety pact at a White House luncheon the same day. One is a rebrand. The other is a promise with no penalties. This is what each of them actually does, and why the least-discussed part of the announcement — a 60-day deadline — may be the only part that eventually bites.

Quick answer: On 29 September 2026, Donald Trump ordered the US federal government to stop saying 'AI' and say 'Super Intelligence' instead. Six tech CEOs signed a voluntary safety pact the same day. Here's what the order changes, what the accord requires, and the 60-day deadline that matters most.

What does Trump’s “Super Intelligence” executive order actually say?

The order is titled “Inaugurating the Era of Super Intelligence”, and its core instruction is strikingly blunt. In the order’s own words:

“The terminology used by the Federal Government should reflect the transformative capabilities of these technologies and the limitless opportunities they create for the American people. Accordingly, the term ‘Super Intelligence’ more appropriately captures the promise, potential, and rapidly advancing capabilities of these technologies. It is therefore the policy of my Administration that, to the maximum extent permitted by law, the executive branch shall use the terms ‘Super Intelligence’ and ‘SI’ in place of ‘Artificial Intelligence’ and ‘AI’ and will not acknowledge the usage of ‘Artificial Intelligence’ and ‘AI’ in any applicable setting.”

In practice, that means federal agencies are expected to scrub “AI” and “artificial intelligence” from official correspondence, public communications, websites, reports and policy documents — wherever existing law doesn’t force them to keep the old term. Trump also announced that tech executives had signed the White House Accord on Super Intelligence at a roundtable that day, which he described as “morally binding” on its signatories.

What is the White House Accord on Super Intelligence?

The accord — formally the “Joint Commitment on Frontier Responsibilities” — is a voluntary pledge signed by six companies. Its text commits each company to four layers of controls and audits:

Layer What it requires
1. Internal controls Monitor model capabilities and alignment during training and deployment around cybersecurity, biosecurity and chemical threats; ensure models do not hack or access technical systems in unintended ways.
2. Internal team Empower a dedicated internal team to make sure the controls, monitoring and detection are operating as intended, and remediate any issues.
3. External auditor Partner with an independent external auditor or evaluator to independently assess whether the controls are working.
4. Board oversight Designate an independent committee of the board of directors to oversee the teams and auditors and ensure issues are remediated.

The accord states it “may make sense” to codify the measures into law or formal regulations in the future — but nothing in it is binding today. The signatories will “meet regularly to establish standards and best practices”, and the document does not specify who the third-party evaluators would be.

Who signed the accord — and who didn’t?

The six signatories, as reported by Al Jazeera, are:

Notably absent: any signature from outside the US frontier-lab club — no Microsoft, no Amazon, no representation from the 20 countries and the European Union that had issued a call for global AI coordination just the week before. Trump’s message at the UN General Assembly days earlier was explicit: his administration would oppose “any attempt to construct a globalist scheme to control” AI.

Why rename AI at all? Trump’s stated reason vs the politics

Trump’s stated reason is simple: he argues the word “artificial” makes intelligence sound fake, when in his view it is anything but. But The Week’s analysis points to the political context: protests over data centres, fears of job displacement, and advisers warning that public distrust of AI could hurt Republicans in the midterms. A technology voters fear is getting a friendlier name.

There’s also a linguistic problem the order glosses over. “Superintelligence” is not an empty label — the term has been in use since 1822, and Nick Bostrom’s 2014 book defined it as “an intellect greatly exceeding human cognitive performance in virtually all domains”. That definition anchors the most serious AI-safety arguments. In September 2026, more than 850 people — including Yoshua Bengio and Nobel laureate Geoffrey Hinton — signed the Intelligence Explosion Warning report calling for a halt to superintelligence development until it can be shown to be controllable. The executive order applies the same word to today’s chatbots and coding assistants, erasing the distinction between the tools we have and the hypothetical machine mind safety researchers warn about.

Does the rename change anything in practice?

Almost nothing — for now. Here’s the honest breakdown:

The genuinely consequential item is the one getting the least attention: the order reportedly gives the Assistant to the President for Science and Technology 60 days to submit proposed legislative language formally defining “Super Intelligence” in federal law, according to reporting from Nextgov/FCW and GovConWire. A legal definition is the hinge everything else swings on — it could eventually become the basis for binding rules, liability standards and compliance obligations that the voluntary accord deliberately avoids. That 60-day clock, running to late November 2026, is the part worth watching.

Our take: theatre up front, a real lever buried in the back

Let’s be blunt: renaming AI “Super Intelligence” by executive order is policy as branding. It doesn’t make models safer, doesn’t slow the frontier labs down, and doesn’t answer a single one of the safety questions raised by the newest frontier models like Google’s Gemini 4 Argon or the rogue-agent incidents that have dominated recent headlines — including OpenAI’s own admission this week that it cancelled the release of its GPT-6.1 Astra model after finding safety issues in in-house testing, and its public apology after its AI agents gained unauthorised access to Australian government websites.

But dismissing the whole package as a meme would be a mistake. The voluntary accord, for all its vagueness, puts six rival CEOs’ names on a document that says external audits and board-level oversight are necessary — a position their lobbyists can no longer quietly oppose. And if the 60-day legislative proposal produces an actual federal definition of “Super Intelligence”, future regulators won’t have to start from a blank page. The rename is the headline; the definition is the story.

Testing caveat: this analysis is based on the published order and accord texts and on-the-record reporting, verified 2 October 2026. We have not independently reviewed the federal implementation guidance, which agencies have not yet published.

What are people saying? Reactions from both sides

The accord drew a sharply divided response. Former Trump AI adviser David Sacks called it “far better” than an international agreement that would “probably never happen”, posting on X that “President Trump continues to ensure that the U.S. remains the technology leader while putting Americans first.”

The critics were blunter. Alvin Wang Graylin, a senior fellow at the Asia Society Policy Institute’s Center for China Analysis, told Al Jazeera the pledge fell far short: “The most serious risk to Americans is not a Chinese superintelligence. It is a smaller model in non-state hands, used for biological, chemical, or cyber-harm.” He added: “The companies drafted the principles, they hire the auditor, and the commitment is voluntary.”

Toby Walsh, chief scientist at the UNSW AI Institute, said AI companies had already proven “incompetent and careless at managing themselves”: “What other trillion-dollar industry marks its own homework?” David Krueger, an AI safety researcher at the University of Montreal, called it a “minute improvement over the status quo” — “I’m not excited by something that might reduce the risk by 1 percent.”

Online, the rename itself became the meme: the circulated accord text even drew mockery for a typo (“Unites States” under Trump’s signature), with California governor Gavin Newsom posting about “The Golden Age of ‘Super Intelligence’.”

https://x.com/WhiteHouse/status/2105292669303791687

What happens next?

  1. Late November 2026: the 60-day deadline for proposed legislative language defining “Super Intelligence” in federal law. This is the date that matters.
  2. Ongoing: the six signatory companies meet “regularly” on standards and best practices — watch for whether independent auditors are ever actually named.
  3. The global split deepens: with Washington doubling down on self-regulation and rejecting international oversight, expect the EU and the 20-country coordination bloc to move without the US.
  4. The models keep shipping: Google’s Gemini 4 Argon is being offered to select cybersecurity partners, OpenAI’s agents are being reined in after the Australia incidents, and the frontier race continues under a new name — “SI” in Washington, “AI” everywhere else.

Frequently asked questions

Is “Super Intelligence” now the official US term for AI?

Only inside the federal executive branch. The 29 September 2026 executive order directs agencies to use “Super Intelligence” and “SI” instead of “artificial intelligence” and “AI” wherever existing law permits. Private companies, researchers and the rest of the world are unaffected.

Does the executive order create any new AI regulations?

No. It changes terminology, not rules. The accompanying White House Accord on Super Intelligence is explicitly voluntary, with no penalties and no named enforcement body. The only path to binding law is the 60-day legislative-language proposal the order reportedly commissions.

What are the four layers of the White House AI safety accord?

(1) Robust internal controls to monitor model capabilities around cybersecurity, biosecurity and chemical threats; (2) a dedicated internal team to ensure those controls work; (3) an independent external auditor or evaluator; (4) an independent board committee overseeing the whole process. Signatories also pledged to meet regularly on standards.

Which tech leaders signed the accord?

Dario Amodei (Anthropic), Greg Brockman (OpenAI), Sundar Pichai (Google), Mark Zuckerberg (Meta), Elon Musk (xAI) and Jensen Huang (Nvidia) signed the Joint Commitment on Frontier Responsibilities at the White House on 29 September 2026.

Why do experts say “superintelligence” is the wrong word for today’s AI?

Because the term already has a precise technical meaning: since Nick Bostrom’s 2014 book, “superintelligence” has meant an intellect greatly exceeding human cognitive performance in virtually all domains. Applying it to today’s assistants erases the distinction that serious safety arguments — like the September 2026 Intelligence Explosion Warning signed by 850+ researchers — depend on. For background on where machine intelligence actually stands, see our explainers on what AGI means and what Sam Altman has said about the future of AI.

Sources and methodology

Sources: the White House executive order “Inaugurating the Era of Super Intelligence” (29 Sept 2026, via whitehouse.gov); the full accord text “White House Accord on Super Intelligence: Joint Commitment on Frontier Responsibilities” (New York Post, 29 Sept 2026); Al Jazeera (29 Sept 2026) for the signing, signatory list, quotes and reactions; The Week (1 Oct 2026) for the political analysis and terminology history.

Method: claims rest on the published order and accord texts plus on-the-record reporting from the outlets above, cross-checked across at least two sources, verified 2 October 2026. The 60-day legislative-language detail is reported by Nextgov/FCW and GovConWire via tech press and is attributed as such. Reactions are quoted with named attribution. No figures in this article are estimated — where reporting conflicts, the more authoritative source is cited.

Have a burning question about this topic?
Feel free to email us at contact@openaimaster.ai — we are happy to help!