A month into Trump’s “Super Intelligence” order, the term lives mainly in agency paperwork

At the end of September, President Trump ordered federal agencies to replace “artificial intelligence” with “Super Intelligence,” after an X poll offering alternatives including “Extreme Intelligence” and “Supreme Intelligence.” The same day, six tech leaders signed a nonbinding White House Accord on Super Intelligence (the accord, the rename). MIT Technology Review checked who has followed. Agencies are swapping the term into websites and solicitations, including an ICE request for “SI-powered” surveillance at detention facilities, but adoption is patchy even inside government, and the order doesn’t apply to previously issued regulations or documents.

Signal Desk will keep saying AI, because it is the accurate term for what exists today. Calling current systems superintelligent overstates what they can do, inflating expectations and blurring the safety debate. When “superintelligence” also names a hypothetical future risk, applying it to today’s chatbots makes both conversations harder. Branding has been critiqued as a substitute for governance since the rename (AI’s image problem), and it doesn’t change consumer economics either (2% paying).

Past renamings became loyalty signals; the administration barred the AP from some White House events over one. If that happens here, the word someone uses for AI will signal their politics before they say anything about the technology. AI is one of the few big issues where Americans don’t yet split cleanly by party, with disagreements on jobs, mental health and warfare cutting across lines. That is worth protecting, especially with voters telling pollsters AI matters to their midterm choices. There is a practical cost too: a legal definition written to fit a brand rather than the technology could shape procurement, oversight and future rules in unintended ways. More in our AI coverage and regulation coverage.