Hours after unveiling the voluntary pact with executives from OpenAI, Google, Meta, Anthropic, Nvidia and xAI, President Donald Trump confirmed the federal executive will effectively rename the technology. Per Euronews (30 September 2026), he signed an executive order telling the government to use “super intelligence” and the acronym “SI” whenever possible.
The order, as quoted by Euronews, says “Super Intelligence” more appropriately captures “the promise, potential and rapidly advancing capabilities” of these technologies. In the preceding press appearance, Trump had told reporters that the word “artificial” sounds bad and that “super is the best word of all,” according to The Verge.
The change does not by itself amend Acts of Congress or the mandates of agencies such as the FTC or the Justice Department. It does force official vocabulary in releases, RFIs, contracts and Administration materials, and pushes media and contractors to decide whether to adopt the new label or keep “AI.”
Timing is deliberate: the White House lunch convened 31 guests including Zuckerberg, Musk, Bezos, Nadella and Pichai. The political message binds three pieces: “morally binding” self-regulation, leadership versus China, and a linguistic rebrand that presents the technology as desirable power rather than a suspicious artificial artefact.
Critics of the approach see a risk of confusion with the technical term “artificial superintelligence” (ASI) used by existential-risk researchers and bills such as Sanders and Casar’s Ban Artificial Superintelligence Act. Blending presidential marketing with a beyond-human capability concept can dilute scientific warnings.
The FTC, according to the New York Post on 30 September, already speaks internally of an “SI race” when describing its probe of frontier labs. Official language may seep into CIDs, testimony and press notes, aligning consumer oversight with a geopolitical-race narrative.
For labs and law firms, the practical impact is documentary: update compliance glossaries, check whether government contracts must reflect “SI,” and clarify in prospectuses (such as Anthropic’s) that the executive’s “super intelligence” is not automatically the ASI thresholds defined in safety literature.
Internationally, Brussels and London still say “artificial intelligence” in the AI Act and UK frameworks. U.S.–EU semantic divergence complicates treaties, technical annexes and official translations: the same system may be called GPAI in Europe and SI in Washington memos.
The rebrand also interacts with the NPR/PBS/Marist poll cited by Euronews: 65% of registered voters oppose data centres in their own communities. Calling the technology “super” does not fix electricity bills or local opposition, but it tries to shift the emotional frame of the midterm debate.
Marketing teams and tech media already face an editorial choice: adopt SI in U.S. headlines while keeping AI for LatAm/Europe? Neural Pulse will keep using “artificial intelligence” in Spanish and English as the standard descriptive term, and will cite “super intelligence” when referring to the order or to beyond-human capability thresholds.
In short, the 29 September order is cheap to sign and costly to ignore in federal paperwork. It creates neither audits nor training caps; it redefines the language with which the Administration wants the public and the bureaucracy to think about the technology race.
The political close is clear: Trump offers CEOs a pact without penalties and, at the same time, a state slogan. Those who want laws with teeth—Khanna, Sanders, or Europe’s AI Office—must fight not only over rules, but over the dictionary that tells the story.
