At the end of September President Trump signed an executive order instructing federal agencies to stop using “artificial intelligence” or “AI” and instead adopt the term “Super Intelligence” (abbreviated “SI”). The administration’s stated rationale is that current systems amplify human ingenuity rather than merely imitate intelligence, and official terminology should reflect that distinction.
Where the new term has appeared
The change is rolling out unevenly. “Super Intelligence” has begun to show up in government solicitations: Immigration and Customs Enforcement (ICE) issued a new request asking companies to build “SI-powered” surveillance for detention facilities. Many government websites have quietly swapped the terminology; some, including the National Institutes of Health (NIH), added banners warning visitors they may encounter both the old and new terms during the transition.
There have been visible stumbles. The day after the rebrand order, the Pentagon announced Project Meridian, a major initiative to study the future of warfare, using the older terminology. A Federal Trade Commission press release used “superintelligence” as one lowercase word rather than the two-word, capitalized form the president ordered. Recent press releases from the Department of Health and Human Services show little sign of the new language.
Given the proliferation of AI-related offices, initiatives, and program names in recent years—such as the Pentagon’s Chief Digital and Artificial Intelligence Office—the abbreviation “AI” is likely to persist. The executive order does not apply retroactively to previously issued regulations or documents.
Adoption among government officials
Government employees appear to be slow to shift from “AI” to “SI.” In interviews a national lab researcher repeatedly said “AI” before acknowledging the slip; Defense Secretary Pete Hegseth used “AI” in an interview on October 3.
Media, politics and public messaging
Past Trump rebrands suggest the administration aims not only to change government vocabulary but to shape public speech as well. On October 27 the president criticized the Wall Street Journal for not using “Super Intelligence,” claiming it is “the name which virtually everyone else is using, or soon will be,” and hinting that future rules might go beyond the executive order to restrict use of “artificial intelligence” in government documents. The Wall Street Journal is not alone in resisting the change; outlets like Newsmax also continue to use the older term.
Tech industry response
On the same day the order was signed, the White House published the nonbinding White House Accord on Super Intelligence, signed by six major tech leaders: Sundar Pichai (Google), Dario Amodei (Anthropic), Mark Zuckerberg (Meta), Greg Brockman (OpenAI), Elon Musk (xAI), and Jensen Huang (Nvidia). Even among those signees, however, uptake of the new phrase is inconsistent.
An analysis of third-quarter earnings calls found that only Mark Zuckerberg and Elon Musk used the term consistently; Meta had already been using “superintelligence” prior to the order. Musk’s companies appear to use the term most often — he even pledged to rename SpaceXAI to SpaceXSI — but Musk himself alternates between SI and AI in posts on X. Executives from Amazon, Microsoft and Alphabet generally used “AI” on earnings calls, though Microsoft’s recent letter to shareholders referred to the “Super Intelligence era.” Jensen Huang has mentioned the term while still using “AI” as well.
Administrative costs and political risks
Critics note the rebrand will consume time and taxpayer dollars. Corporate leaders can choose their internal branding, but government staff are preparing guidance documents to navigate the transition. The executive order requires leaders of federal agencies to collaborate on drafting a legislative definition of “Super Intelligence” to be submitted to the president by the end of the month.
Beyond administrative expense, the more consequential risk is to public discourse. Informal conversations about AI have become common in everyday settings; while many of those discussions focus on fears—job losses or AI-enabled warfare—people also frequently express a desire to have a say in how the technology shapes their lives. Currently, attitudes toward AI do not line up neatly with partisan divisions; that makes cross-aisle conversation easier than on some other polarizing topics.
If the term “Super Intelligence” turns into a political loyalty test, that may change. It is easy to imagine discussions between strangers ending as soon as one uses “super intelligence” and the other uses “AI.” Such a split would undercut the broad-based conversation that many see as essential while the country is only beginning to grapple with AI’s societal impact.
Conclusion
One month after the executive order, rollout of “Super Intelligence” is partial and inconsistent across government and industry. The administrative burden and required legal definitions are immediate consequences, but the longer-term concern is political: turning a technical term into a marker of political allegiance could fray the inclusive public debate about AI at a time when that conversation is only getting started.
(Note: the original analysis appeared in The Algorithm newsletter and assessed developments based on public statements and documents.)
