November is fully booked! Next classes blast off the week of Mon, Dec 7.

Reserve your spot
Super Intelligence Policy: Why the White House Changed Words
Table of Contents

Super Intelligence Policy: Why the White House Changed Words

A September 2026 executive order renamed AI as super intelligence. What the super intelligence policy shift actually changes, and what it deliberately does not.

The federal government renamed artificial intelligence and kept the definition. That is the whole story of the new super intelligence policy in one sentence, and it is more interesting than either the cheerleading or the mockery suggests. Executive Order 14434, signed September 29, 2026, tells executive-branch agencies to write “Super Intelligence” and “SI” in their correspondence, websites, reports and policy documents. Then it defines the term by pointing at the existing statutory definition of artificial intelligence in federal law. Same scope, new label. And one deadline buried near the end that does the actual work.

Key Takeaways

  • Executive Order 14434, dated September 29, 2026, directs executive departments to use “Super Intelligence” and “SI” rather than “artificial intelligence” and “AI,” “to the maximum extent permitted by law.”
  • The order defines SI as “the technologies and systems encompassed by the term ‘artificial intelligence’” under section 9401(3) of title 15, United States Code. Nothing about what counts as covered technology changed.
  • The operative provision is a 60-day deadline for the Assistant to the President for Science and Technology to submit proposed legislative language for a federal SI definition. A statutory definition would change scope; the renaming alone does not.
  • A separate announcement on October 4, 2026 established a Super Intelligence Force, chaired by National Intelligence Director Jay Clayton, with 120 days to report on risks and opportunities.
  • For families and schools, nothing enforceable changed. What changed is the vocabulary that the next round of bills, grant applications and district policies will be written in.

What the order actually says

An executive order is an instruction from the president to the executive branch. It can direct agencies. It cannot by itself bind a state, a school district, a company or a parent.

The order, titled “Inaugurating The Era Of Super Intelligence,” makes its reasoning explicit: frontier systems now do “much more than imitate or automate discrete aspects of human intelligence,” and so “the terminology used by the Federal Government should reflect the transformative capabilities of these technologies and the limitless opportunities they create.”

Then comes the part worth reading twice. The order defines super intelligence as the technologies already covered by the statutory term. That statute, 15 U.S.C. § 9401(3), part of the National Artificial Intelligence Initiative Act of 2020, defines artificial intelligence as “a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.”

Read that definition carefully, because it is remarkably wide. A credit-scoring model fits. A spam filter fits. The recommendation engine on a video app fits. A thermostat with a learned schedule arguably fits. The statutory definition was never about human-level cognition, and the new label does not narrow it to things that are actually superintelligent in any technical sense.

So the honest summary is: the executive branch now calls a very broad category of software by a very dramatic name, while the legal boundary of that category is unchanged.

One correction to the record. Coverage, including the summary our own research notes started from, placed the rebrand on October 4. The executive order is dated September 29, 2026. TechCrunch’s video segment on the terminology change ran October 1. The October 4 date belongs to the separate task-force announcement described below.

Why super intelligence policy starts with a definition

In regulation, definitions are the thing. They decide who is covered, which agency has jurisdiction, what has to be reported, and what triggers a penalty. Everything downstream is detail.

The clearest example is in Europe. The EU AI Act implementation timeline is a sequence of dates attached to defined categories. The Act entered into force on 1 August 2024. Prohibitions and AI-literacy duties started applying on 2 February 2025. Rules for general-purpose AI models started on 2 August 2025, with providers of models already on the market given until 2 August 2027. Obligations for high-risk systems under Annex III begin on 2 December 2027, and under Annex I on 2 August 2028. Whether your product is a “general-purpose AI model” or a “high-risk system” determines which of those dates applies to you, and the difference is worth millions to a company.

The United States has nothing equivalent in force. What it has is the voluntary NIST AI Risk Management Framework, released January 26, 2023 and organised around four functions: govern, map, measure and manage. Its Generative AI Profile, NIST AI 600-1, arrived July 26, 2024. Agencies cite it. Nobody is fined for ignoring it.

Which is why the 60-day legislative-language deadline in the order matters more than the renaming. Asking the White House science adviser to draft a statutory definition of “super intelligence” is a request to redraw the boundary. A new definition could be narrower than § 9401(3), capturing only frontier models and leaving ordinary machine learning outside. It could be broader. Either way, that draft is the document to read when it surfaces, and it will get roughly one percent of the attention the name change got.

What each instrument can and cannot do

Parents reading AI news are handed a stream of announcements with no indication of which ones carry force. Here is the hierarchy, from weakest to strongest.

InstrumentWho issues itWhat it can actually compelThis month’s example
Public announcement or social postAnyoneNothing. It signals intent.The October 4 Super Intelligence Force announcement
Open letterSignatoriesNothing. It creates a public record of who said what.The “Pacing the Frontier” letter, July 28, 2026
Executive orderPresident, to the executive branchAgency behaviour: terminology, procurement, internal processExecutive Order 14434
Task force reportThe task forceRecommendations. No legal effect on its own.The Force’s 120-day report
Agency ruleA regulator, after notice and commentBinds regulated parties; enforceable with penaltiesNone yet for general-purpose AI in the US
StatuteCongressBinds everyone within its scope15 U.S.C. § 9401 (definitions only)

Most AI news lives in the top three rows. Our companion piece on why an open letter is not the same as regulation works through that distinction with the July letter as the case study.

The Super Intelligence Force, and what a 120-day report usually produces

On October 4, 2026, TechCrunch reported the announcement of a Super Intelligence Force, made in a post on Truth Social. The stated purpose is to “coordinate the effort of the Federal Government to ensure that America continues to lead the World in Super Intelligence.”

The membership is the informative part. National Intelligence Director Jay Clayton chairs it. The vice chairs are FTC Chair Andrew Ferguson, Emil Michael as Undersecretary of War for Research and Engineering, and Scott Kupor at the Office of Personnel Management. That is intelligence, competition enforcement, defence research and federal hiring. Notably absent from the leadership: the Department of Education, and any agency whose primary remit is children.

The group has 120 days to produce a report on risks and opportunities. Its charter, per the reporting, emphasises developing response plans for threats while “preventing overregulation and regulatory capture that would stifle innovation and competition.”

A 120-day interagency report is a well-worn instrument. It typically produces a document, a set of recommendations, and occasionally a follow-on executive order. It does not produce enforceable rules. If you want to predict its output, read the charter language: a body told to prevent overregulation while assessing risk will usually recommend voluntary mechanisms, information sharing and procurement preferences.

What this means in a house with kids

Nothing enforceable. I want to be direct about that, because the alternative is pretending that a vocabulary change in federal correspondence affects your eleven-year-old this winter.

What it does affect is language, and language eventually reaches your school.

Expect the term to show up in district paperwork

Federal grant programmes, procurement language and agency guidance filter into district policy documents, often by copy-paste. If your school’s AI policy is revised in 2027 and suddenly says “SI,” that is where it came from. The substance of a policy matters more than its nouns, so when you read one, look for the three things that are actually operational: who may use what, what counts as cheating, and what the school does with student data.

Teach the word’s real history, not the press release

“Superintelligence” is a term with a specific scholarly lineage, not a marketing word invented in 2026. Nick Bostrom’s Superintelligence: Paths, Dangers, Strategies, published by Oxford University Press in 2014, popularised it to describe a hypothetical system that exceeds human cognitive performance across essentially all domains. Current systems do not meet that description, and the executive order does not claim they do; it points at a statutory definition covering ordinary prediction systems. A kid who knows the difference between a technical term and its political adoption has a skill that generalises well beyond this story. Our piece on what to actually worry about with superintelligence separates the research question from the headlines.

Use this as the cleanest possible lesson in reading a primary source

Executive orders are short, free, and written in plain-ish English. This one runs a few paragraphs. Open it with a twelve-year-old, find the sentence that defines the term, follow the citation to the statute, and read that definition too. Twenty minutes. The lesson is not about AI. It is that the actual document is usually available and usually says something slightly different from the coverage.

Watch the definition, not the name

When the proposed legislative language appears, the only question that matters is scope: does the new definition cover frontier models only, or everything that makes a prediction? That single choice determines whether future rules touch the tools in your child’s classroom or only the labs in California.

What not to do

Do not argue about the word. It is the most available thing to argue about and the least consequential. People will spend a year on whether “super intelligence” is accurate or absurd while the legislative definition gets drafted with almost no public attention. Spend your attention on the second thing.

What to Watch For Over the Next 3 Months

  • Week 4: The 60-day legislative-language deadline falls in late November 2026. Watch for whether proposed statutory text becomes public, and whether its definition is narrower or broader than 15 U.S.C. § 9401(3).
  • Month 2 red flags: Watch whether federal education guidance adopts the new terminology without changing any requirement. A document that renames and does nothing else is a signal that the substantive policy work has stalled.
  • Month 3 self-check: The Super Intelligence Force’s 120-day report comes due around early February 2027. When it lands, skim for one thing only: does it recommend any mechanism that can be enforced against a company, or only mechanisms companies opt into?

Frequently Asked Questions

Did the government decide that AI is now superintelligent?

No. The order changes what federal agencies call the technology and explicitly ties the new term to the existing statutory definition of artificial intelligence, which covers any machine-based system that makes predictions, recommendations or decisions for human-defined objectives. That definition includes a spam filter.

Does this executive order regulate AI companies?

No. An executive order directs the executive branch. It can change how agencies write, buy and report. Binding obligations on companies require either a statute from Congress or a rule made by an agency through notice-and-comment procedures, and neither exists for general-purpose AI in the US as of early 2027.

What is the Super Intelligence Force going to do?

Produce a report within 120 days on risks and opportunities, chaired by the National Intelligence Director with vice chairs from the FTC, defence research and federal personnel management. Its charter asks it to plan responses to threats while avoiding what it calls overregulation and regulatory capture.

Does any of this change school AI rules?

Not directly, and not this year. District rules come from state law, school boards and procurement contracts. Federal terminology tends to arrive in district documents a year or two later through grant language and agency guidance.

Why do definitions matter so much in AI law?

Because they set scope. The EU AI Act attaches different deadlines and duties to defined categories, so whether a system is “high-risk” or a “general-purpose AI model” determines which obligations apply and when. A fight over a definition is a fight over who is regulated.

Is “superintelligence” a real technical concept?

Yes, as a hypothetical. It describes a system that would exceed human performance across essentially all cognitive domains, a framing popularised by Nick Bostrom’s 2014 Oxford University Press book. No deployed system meets that description, and the executive order does not assert otherwise.


About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.


Sources

  1. The White House. (2026). “Inaugurating The Era Of Super Intelligence.” Executive Order 14434, September 29, 2026. https://www.whitehouse.gov/presidential-actions/2026/09/inaugurating-the-era-of-super-intelligence/
  2. Legal Information Institute, Cornell Law School. “15 U.S. Code § 9401 — Definitions,” National Artificial Intelligence Initiative Act of 2020. https://www.law.cornell.edu/uscode/text/15/9401
  3. Ha, Anthony. (2026). “Trump unveils his new Super Intelligence Force.” TechCrunch, October 4, 2026. https://techcrunch.com/2026/10/04/trump-unveils-his-new-super-intelligence-force/
  4. Loconsolo, Theresa. (2026). “It’s not AI anymore, it’s ‘super intelligence’ (according to the White House).” TechCrunch, October 1, 2026. https://techcrunch.com/video/its-not-ai-anymore-its-super-intelligence-according-to-the-white-house/
  5. Future of Life Institute. “EU AI Act Implementation Timeline.” https://artificialintelligenceact.eu/implementation-timeline/
  6. National Institute of Standards and Technology. (2023). “AI Risk Management Framework.” https://www.nist.gov/itl/ai-risk-management-framework
  7. Bostrom, Nick. (2014). Superintelligence: Paths, Dangers, Strategies. Oxford University Press.
Ricky Flores
Written by Ricky Flores

Founder of HiWave Makers and electrical engineer with 15+ years working on projects with Apple, Samsung, Texas Instruments, and other Fortune 500 companies. He writes about how kids learn to build, think, and create in a tech-driven world.