Introduction
On September 29, 2026, President Donald Trump signed Executive Order 14434, titled "Inaugurating the Era of Super Intelligence," directing the entire federal executive branch to replace the terms "artificial intelligence" and "AI" with "Super Intelligence" and "SI." He had previewed the switch a week earlier at the United Nations General Assembly. Within days, headlines called it a rebrand, a distraction, and a declaration of ambition — sometimes all three.
This post explains what the order actually says, what it changes, what it deliberately does not change, and why the wording matters for the ASI debate even though the technology is untouched. (For the timeline forecasts behind that debate, see Post 4.)
Quick Takeaways
- Executive Order 14434 (September 29, 2026) orders federal agencies to use "Super Intelligence"/"SI" instead of "artificial intelligence"/"AI" in official communications.
- The order is about vocabulary, not capability: it explicitly defines "Super Intelligence" by pointing at the existing legal definition of AI (15 U.S.C. 9401(3)).
- The same day, six tech leaders — Amodei, Pichai, Zuckerberg, Brockman, Huang, Musk — signed a voluntary White House accord on frontier-model commitments.
- The Assistant to the President for Science and Technology has 60 days to propose a formal federal definition of SI.
- What it does not do: change any law, regulation, contract or grant; earlier records are exempt.
What the Order Says
Section 1 sets out the rationale: American innovators are pushing the frontier, today's systems "increasingly amplify human ingenuity and unlock new forms of creativity," and — the key sentence — these capabilities "increasingly represent not merely artificial intelligence, but a new era of Super Intelligence." The operative directive follows: to the maximum extent permitted by law, the executive branch shall use "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI," and will not acknowledge the old usage, in official correspondence, public communications, policy documents and other non-statutory materials.
Then comes the part most commentary skipped. Section 3(a) defines the new term — and defines it by reference to the old one. For the purposes of the order, "Super Intelligence" means the technologies covered by the existing statutory definition of artificial intelligence in 15 U.S.C. 9401(3), from the National Artificial Intelligence Initiative Act of 2020. That definition is broad: roughly, any machine-based system that makes predictions, recommendations or decisions influencing real or virtual environments. On a plain reading, a spam filter qualifies. Nothing in it requires human-level performance.
So the order makes two claims at once: a rhetorical claim (these technologies represent a new era) and a legal definition that changes nothing. The name changed overnight. The statutory concept did not.
The Accord and the Context
The signing ceremony included a second document: a voluntary accord, described in press reports as a "Joint Commitment on Frontier Responsibilities." Six technology leaders signed it — Dario Amodei (Anthropic), Sundar Pichai (Google), Mark Zuckerberg (Meta), Greg Brockman (OpenAI), Jensen Huang (Nvidia) and Elon Musk (xAI), per Associated Press reporting. The accord commits each company to monitoring its models, maintaining internal oversight, using outside evaluators, and establishing board-level committees. President Trump called it "morally binding"; commentators on TechCrunch's Equity podcast called it "deeply non-binding" and framed the day as a rebranding exercise.
The context matters too. In the same week, the administration launched an AI-powered front door for federal services — and, according to reporting by Reuters and The New York Times, the Federal Trade Commission opened or expanded investigations into frontier labs including Anthropic and OpenAI. The voluntary handshake and the regulatory scrutiny happened simultaneously, which says something about how seriously the "new era" framing describes the government's actual posture.
Why the Wording Matters Anyway
If the order changes nothing technical or legal, why does this site cover it? Three reasons.
First, language shapes the debate's boundaries. For years, "superintelligence" was the word the ban letter (Post 14) used for the thing to be prohibited — a boundary marker for a hypothetical to be blocked, not branded. The executive order takes the same word and makes it the official name for today's technology. The term now means two incompatible things: the hypothetical endpoint the Future of Life Institute wants banned, and the current product line the White House wants celebrated. When one word means both "stop this" and "we lead at this," public conversation gets harder, not easier.
Second, it is a claim about capability disguised as a claim about vocabulary. The order asserts that today's systems "increasingly represent" a new era of Super Intelligence — a statement about what the technology can do, presented as a reason for renaming it. But the order's own definition requires no human-level performance. Readers should separate the two: the renaming is real and official; the capability claim is rhetoric, and this site evaluates capability claims against evidence, not proclamations — including what researchers actually say about jobs and the economy in Post 18.
Third, the 60-day clock is the part to watch. The order tasks the President's science adviser with proposing a formal federal definition of "Super Intelligence," with legislative language, within 60 days — so, by late November 2026. If that definition narrows the term to genuinely frontier capabilities, the rebrand gains substance. If it stays as broad as the current pointer-definition, "SI" will officially mean "everything from spam filters up," and the era inaugurated will be an era of confusion. ASI News World will report the definition when it lands.
One concrete sign the rename is already happening: the federal body that evaluates advanced models — created at NIST as the U.S. AI Safety Institute, then renamed the Center for AI Standards and Innovation in June 2025 — is now described on NIST's website as the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), its third name in roughly three years, per a Cloud Security Alliance research note. The center's stated mission still includes evaluating demonstrable risks in cybersecurity, biosecurity and chemical weapons. The name changes; the evaluation work continues.
The Bottom Line
Executive Order 14434 is a real, official renaming with no technical or legal substance — yet. It matters because it conscripts the word "superintelligence" into government branding at the exact moment others use it to demand prohibition, and because the promised November definition will reveal whether the administration means the word or just likes the sound of it. Not proven by this order: that today's systems are superintelligent. Proven: that the fight over the word is now a fight over the future, conducted in public.




0 Response to ""Super Intelligence" as a US Government Term: The September 2026 Executive Order, Explained"
Post a Comment