Trump rebrands AI as 'superintelligence' in the US and rejects global rules for its development

By Canal 1 · September 22, 2026.

Donald Trump announced before the UN General Assembly that the US Administration will stop using the term "artificial intelligence" in its official documentation. In its place, it will adopt "superintelligence": according to El Sol de México's account of the same speech, the official acronym will shift from AI to SI. The justification Trump himself gave was linguistic: the word "artificial" suggests something fake or inauthentic, and that, he said, does not reflect the real scope the technology has acquired.

It is not a technical or universal change. Artificial intelligence will still be called that in scientific, technological and academic circles; what changes is the vocabulary the US Government will use in its own paperwork. But the announcement did not come in isolation: in the same speech, Trump rejected the possibility of an international framework to control AI development and argued that the US must retain wide latitude to keep advancing unchecked, in what he described as a strategic race against China. It also comes at the same General Assembly where — according to other reporting by Canal 1 — leaders of 22 countries were calling for precisely that: binding safety measures for AI; a request this speech effectively dismisses.

The two things are worth separating, because they carry very different weight. Changing a word on a letterhead does not move a single line of code or a single benchmark: AI systems in 2026 are no more or less capable for being called "superintelligence" in a US Government document. Rejecting international governance, by contrast, is a decision with real consequences: it sets Washington's position in the dispute, still open within the United Nations, between those demanding shared rules to contain risks and those who prefer each power to run at its own pace.

That is where the name change stops being trivial and becomes revealing as a symptom. "Superintelligence" is not a neutral term: in AI research it denotes a hypothetical threshold at which systems would vastly exceed human capability at practically any task, a scenario no serious lab claims to have reached yet. Adopting it as the official label for today's AI, rather than for a future milestone, means appropriating the sector's most ambitious vocabulary to dress up a present still far below that promise. It is the same rhetorical inflation some companies already practice when they talk about "AGI" without data to back it up: the word runs ahead of demonstrated capability.

That it is the White House, and not a lab or an investor, doing so adds a further nuance. A government that takes for granted that it already wields "superintelligence" can hardly present itself, at the same time, as the prudent actor asking for time to think about safeguards; the language itself pushes toward acceleration, not caution. And it coincides, not by chance, with the refusal to let an international body set conditions on development: if the technology is already "super," slowing it down — or letting other countries limit it — starts to sound, within that rhetorical framing, like nonsense.

Our reading is that this story matters less for what it changes today than for what it foreshadows. A government's official vocabulary is not a cosmetic detail: it shapes which questions seem reasonable and which seem unwarranted alarmism. If in Washington AI comes to be called "superintelligence" in every report and every contract, any subsequent debate about limits, audits or international cooperation will already start from an inflated premise. In the short term this favors the narrative of a race without brakes, and there lies a genuine risk: less room for multilateral cooperation just when the safety of increasingly agentic systems should be a matter for coordination, not semantic competition. In the long term, however, no label changes what really matters: whether these systems end up bringing us closer to curing diseases, extending healthy life and spreading abundance will depend on the real capability that gets built and on how it is governed, not on what it is called on a 2026 letterhead.