Washington officially renamed AI to 'Super Intelligence' — what the executive order actually changes
On September 29, 2026, President Trump signed an executive order directing federal agencies to use 'Super Intelligence' (SI) instead of 'artificial intelligence' in official documents — the same day six major AI company executives signed a voluntary safety accord at the White House. Both are real, and both matter less than their headlines suggest.
- What the order mandates — The executive order directs all executive-branch departments and agencies 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'. The definition is borrowed from section 9401(3) of title 15 of the US Code, so it renames the field without redefining it.
- The day's second headline — At the same White House event, the heads of Anthropic, OpenAI, Google, Meta, xAI and Nvidia signed the 'Joint Commitment on Frontier Responsibilities', pledging internal controls and monitoring teams plus partner audits by independent external auditors. Politico described the pact as 'morally binding'; it does not stipulate who carries out third-party evaluations.
- What stays the same — The order does not touch the models, the labs, or anyone outside the executive branch. In the research field, 'superintelligence' already has a specific technical meaning — systems exceeding human capability across virtually all domains of thought — and it has not been achieved, so the new official name is a stretch before the science.
Once in a while an industry story is about naming, not shipping. On September 29, 2026, President Trump signed an executive order, Inaugurating the Era of Super Intelligence, directing the executive branch to replace "artificial intelligence" with "Super Intelligence" (SI) in official documents, websites and public communications. The White House fact sheet says agencies should "no longer acknowledge" the old terms. The move follows a social-media poll in which Trump asked followers to choose between "SUPER INTELLIGENCE" and "SUPERIOR INTELLIGENCE".
The order was signed September 29, 2026, two days before this article. Everything below comes from the order and White House fact sheet themselves, plus independent reporting from Politico, Al Jazeera and Fox Business published the same day. No federal agency has yet changed its public documents in a way we can point to; compliance is instructed, not yet observable.
What the order actually says
Three elements deserve attention. First, scope: the mandate applies to executive-branch departments and agencies, not to courts, legislatures, states, or private companies. Second, definition: section 3 borrows the existing legal definition of "artificial intelligence" from section 9401(3) of title 15 of the US Code and simply relabels it "Super Intelligence" and "SI". That is a rebranding with a pre-existing legal anchor — it does not create a new regulatory category. Third, process: the order directs the Assistant to the President for Science and Technology to propose a federal definition and identify any further executive action needed to implement it. As of publication, no such proposal has been published.
The order's own justification is aesthetic, not technical. It argues that current systems do "much more than imitate or automate discrete aspects of human intelligence" and that federal terminology "should reflect the transformative capabilities" of the field. The White House fact sheet adds familiar claims: the US led AI development "for over 70 years since its inception as a field of research", more than $5 billion has been announced for the "Genesis Mission", and a National Super Intelligence Policy Framework was unveiled in March 2026.
Why the industry term is doing different work
Here is the friction. In research and policy circles, "superintelligence" has a specific meaning: a system that exceeds the cognitive capabilities of humans in virtually all domains of thinking, including most scientific invention. Politico notes plainly that superintelligent AI "has not yet been achieved". Using the word as a synonym for every model now shipping — from chat assistants to coding agents — doesn't match either the technical term or the order's own borrowed definition. Expect the mismatch to produce ambiguity rather than clarity in the document trail: an agency memo calling a standard LLM workflow "SI" says little more than it called the same thing "AI" yesterday.
The pact signed in the same room
The rename happened at a White House summit that also produced the day's more substantive document: the Joint Commitment on Frontier Responsibilities. Its six signatories are Dario Amodei (Anthropic), Greg Brockman (OpenAI), Sundar Pichai (Google), Mark Zuckerberg (Meta), Elon Musk (xAI) and Jensen Huang (Nvidia). Per Al Jazeera's reporting of the accord, the companies pledge to implement "robust" controls for monitoring AI models, to appoint internal teams responsible for ensuring that controls and detection "are operating as intended, and that any issues are remediated", to partner with independent external auditors, and to meet regularly to establish safety standards and best practices. The accord says it "may make sense" to codify the measures into law or formal regulation in the future.
The criticism is immediate and specific. Toby Walsh of the UNSW AI Institute asked, on the record, "What other trillion-dollar industry marks its own homework?" — pointing out that the companies draft the principles and choose the auditor. David Krueger of the University of Montreal called the document an example of "regulate without regulating", where everything is ostensibly voluntary "but there is an implicit threat of punishment for noncompliance". The accord does not stipulate who carries out third-party evaluations. Context matters too: twenty countries and the EU had proposed a global oversight body just weeks earlier, and Trump told the UN General Assembly that the administration opposes "any attempt to construct a globalist scheme to control" AI.
What this means for you
- For readers of US government documents: expect a two-track vocabulary. Statutes keep saying "artificial intelligence" (definitions in law can't be renamed by an order); agency documents increasingly say "SI". When a policy document switches terms, check which definition it is actually applying — the borrowed one from title 15, not the industry term.
- For AI teams answering procurement or compliance questionnaires: the self-audit pledge becomes a question: "Does your vendor's internal monitoring satisfy the Joint Commitment's 'operating as intended' standard, and who verifies it?" The answer is "the company chooses the auditor" — which is weaker than independent evaluation, and worth flagging in due diligence.
- For product and marketing teams: "super intelligence" is now an official US government label with a borrowed legal definition. Using it in marketing to imply superhuman capability is a different claim than the definition supports; expect tighter scrutiny of that usage, not looser.
What changed: the federal vocabulary, on paper. What didn't: the technology, the safety bar, the regulatory gap, or who gets to audit whom. The rename costs nothing and signals a posture — capability framing over precaution. The voluntary pact is the real document, and its limits are its own text: voluntary, self-selected auditors, no cross-border reach, no independence guarantee. Both artifacts deserve more skepticism than their White House debut warranted.
Bottom line: the executive order is a renaming with a borrowed legal definition, bound to the executive branch, with no new technology content. The Joint Commitment is the substantive artifact — and its weaknesses are documented in its own terms: voluntary, self-policed, unevaluated. Treat "Super Intelligence" in federal documents as "AI" with extra steps, and audit pledges as marketing until an independent evaluator says otherwise.
- White House — Inaugurating the Era of Super Intelligence (Executive Order, September 29, 2026)
- White House — Fact Sheet: President Donald J. Trump Inaugurates the Era of Super Intelligence (September 29, 2026)
- Politico — Trump signs executive order renaming AI to 'super intelligence' (September 29, 2026)
- Al Jazeera — Trump, tech bosses sign voluntary pact pledging 'robust' AI safeguards (September 29, 2026)
- Fox Business — President Trump orders federal agencies to replace AI with 'Super Intelligence' (September 29, 2026)
Prices and benchmark figures are standard published rates as of October 1, 2026 and can change without notice. Some outbound links may be affiliate links — see our disclosure. Found an error? Tell us and we'll correct it in public.