Is AI Now Called Super Intelligence? The SI Order Explained

Is AI Now Officially Called Super Intelligence?

Only inside the federal executive branch. Executive Order 14434, signed September 29, 2026 and published in the Federal Register on October 2, directs executive departments and agencies to use "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI" in their correspondence, public communications, websites, reports, and policy documents. It does not rename anything for businesses, universities, states, or the AI companies themselves.

The change was first announced on September 22, when the President told the United Nations General Assembly that the term would change, saying "the use of the word 'artificial' makes intelligence fake" (AP via Gulf News). The same AP report says State Department staff working on UN affairs were told to change their references before any order existed. The executive order followed a week later, on the same day as a White House meeting with technology leaders.

Search interest moved fast. In Google Trends for the United States (retrieved October 2, 2026), daily interest in "super intelligence" jumped from 3 on September 21 to 53 the day of the UN speech, fell back to 6 within five days, then hit its peak of 100 on September 30, the day after the order was signed. This is a news-driven term, so the useful question for a business is not the label but what, if anything, changed.

Last verified: October 2, 2026, against the Federal Register text of Executive Order 14434. We will update this page when the federal definition required by the order is made public.

What Does Executive Order 14434 Actually Require?

It requires federal agencies to change their vocabulary, "to the maximum extent permitted by law," and nothing more for anyone outside government. Section 2 lists where the new terms apply: official correspondence, public communications, websites, reports, policy documents, and other non-statutory documents within the executive branch. Section 1 goes further in tone, stating that the executive branch "will not acknowledge the usage of 'Artificial Intelligence' and 'AI' in any applicable setting," but it is still scoped to the executive branch.

Three limits in the text matter most:

• Nothing old gets rewritten. Section 2(b) says nothing in the order "requires the alteration of previously issued regulations, Presidential actions, contracts, grants, or other historical documents."
• The legal meaning does not change. Section 3(a) defines "Super Intelligence" and "SI" as the technologies already covered by the statutory definition of artificial intelligence in 15 U.S.C. 9401(3): "a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments."
• A real definition is still coming. Section 3(b) gives the Assistant to the President for Science and Technology 60 days to propose legislative language for a federal definition of Super Intelligence, including whether it should modify or replace the statutory definition of artificial intelligence. The proposal goes to the President, and the order does not require it to be published. Sixty days from September 29 is November 28, 2026.

The White House fact sheet frames the rationale as language: the administration argues "Super Intelligence" better describes what today's systems can do than a term implying imitation of human intelligence.

Is Super Intelligence the Same as Superintelligence?

No, and the spelling is the tell. "Super Intelligence" (two words, abbreviated SI) is the federal government's new name for the AI systems that exist today. "Superintelligence" (one word, often called artificial superintelligence or ASI) is a decades-old research term for a hypothetical system whose intellect greatly exceeds the best human minds in virtually all domains, a framing associated with philosopher Nick Bostrom (Wikipedia).

The two now sit side by side in the news, which is why the distinction matters:

Some companies already used the one-word form before any of this. Meta, for example, announced Meta Superintelligence Labs in June 2025 to pursue that longer-term goal. When you read a headline, the capital S, the space, and the context tell you which meaning is intended. In this article we follow the same rule: Super Intelligence (SI) for the federal term, superintelligence for the research concept.

What Is the White House Accord on Super Intelligence?

It is a voluntary, one-page pledge signed on September 29, 2026 at a White House lunch, separate from the executive order. Its full title is the White House Accord on Super Intelligence: Joint Commitment on Frontier Responsibilities. Reporting by The Next Web and Defense One describes four layers of controls that each signatory company "should" put in place for its frontier models.

The four layers are:

1. Internal controls that monitor model capabilities during training and deployment, with attention to cybersecurity, biosecurity, and chemical threats.
2. An empowered internal team that checks those controls work and that problems get fixed.
3. Independent external auditors or evaluators that assess whether the controls function as intended.
4. An independent board committee that receives the reports and oversees remediation.

The accord uses "should," not "must," and carries no penalties, timeline, or enforcement mechanism. The President called it "morally binding." In the press conference that followed, Meta's Mark Zuckerberg described it as "a start" rather than the only step the industry would take. xAI's Elon Musk summarized it to Defense One as "joint monitoring, board special committees, just generally grading each other's homework."

Which Companies Signed, and Which Did Not?

Six companies signed: Google (Sundar Pichai), Anthropic (Dario Amodei), Meta (Mark Zuckerberg), OpenAI (President Greg Brockman), xAI (Elon Musk), and Nvidia (Jensen Huang), alongside the President. Microsoft and Amazon leaders attended the lunch but their companies did not sign, according to The Next Web, and Yahoo News reported that Apple did not attend.

Note that OpenAI was represented by its president, Greg Brockman, not chief executive Sam Altman.

Are the AI Companies Calling Their Products SI Now?

Not so far. In the press conference that followed the signing, the executives who spoke kept using "AI": Anthropic's Dario Amodei said "whoever wins AI wins," OpenAI's Greg Brockman said "we build artificial intelligence," and Nvidia's Jensen Huang spoke about advancing "the capabilities of AI." One speaker said "AI," then corrected to "SI, pardon me" mid-sentence, which drew laughter. The President, by contrast, said on camera, "We're using the word super because the other word is a fake word."

The order does not ask companies to change, and none of the signatories has relabeled its products. Expect "AI" to remain the market's word, with "SI" appearing wherever you deal directly with federal agencies.

Does the Super Intelligence Order Apply to Private Companies?

No. The executive order binds executive-branch agencies, and the accord binds only its six signatories, voluntarily. Law firm Freshfields notes that the accord "creates no new licensing, preclearance, or mandatory disclosure obligations for AI developers," and IAPP's analysis points out that statutes, regulations, contracts, and state laws are unchanged.

Your existing obligations stay exactly where they were. The EU AI Act and US state AI laws still say "artificial intelligence," and nothing in the order changes them. Our AI compliance deadlines guide and the compliance calendar track those dates.

What Should Businesses Do About the SI Rename?

Nothing urgent, and three practical things. The rename changes vocabulary, not obligations, so the work of adopting AI safely is the same work it was last month.

1. Federal contractors and grant applicants: expect the new term in paperwork. Freshfields advises contractors to expect agency solicitations, forms, and official correspondence to adopt "Super Intelligence" going forward. Map SI to AI in your proposal templates, capability statements, and compliance matrices so a keyword mismatch never costs you an evaluation point.
2. Everyone: watch for the definition. The proposal is due to the President by November 28, though it may not be made public that day. The proposed federal definition could expand or replace the statutory definition of AI. If the President or Congress adopts it, it could change which systems count for future incentives, procurement requirements, and oversight. That is the part of this story most likely to matter for compliance.
3. Treat the accord as a benchmark, not a requirement. The four layers (internal controls, an accountable team, external review, and board oversight) are a reasonable governance pattern for any organization deploying AI, scaled down. If you cannot say who monitors your AI tools today, start with shadow AI: the tools your team already uses without approval.

For definitions of both terms and 30-plus related concepts, see the CoFabrix AI glossary. We also follow lab and industry announcements in AI Pulse. If you want help building the governance layer, our AI Governance & Compliance service scopes it to your size and risk profile.

What Happens Next With Super Intelligence Policy?

The next concrete milestone is the proposed federal definition, due to the President by November 28, 2026, followed by whatever the President or Congress does with it. Until then, "Super Intelligence" in a federal document means exactly what "artificial intelligence" meant before September 29.

Three things are worth watching:

• The definition proposal. Whether it simply renames AI or carves out a new, narrower category for frontier systems will decide whether SI becomes a legal term with obligations attached.
• Accord follow-through. GovConWire reports that the companies committed to regular meetings to develop safety standards. Watch for published audit frameworks or named external evaluators, which would make the voluntary commitments measurable; the accord itself sets no timeline for either.
• Adoption outside government. If labs, standards bodies, or major customers start using SI, the market term could shift. So far, they have not.

Whatever the label, the questions for your organization are unchanged: where you use AI today, what data it touches, and who is accountable for it.

Not sure where your organization's public footprint stands with AI search engines and crawlers? The free Website Health Check scores six categories in under a minute, including AI visibility.