One Month Into Trump's AI Rebrand: Federal Agencies Quietly Swap Terms, Tech Giants Pay Lip Service to the New Name But Rarely Use It Naturally

It has been a full month since Donald Trump renamed AI. The new official name is Super Intelligence, shortened to SI. The only provision in the late September executive order required all federal agencies to refer to what was previously called artificial intelligence as 'Super Intelligence'. This name beat out the other two finalists, Extreme Intelligence and Supreme Intelligence, in a public X poll. It is not unusual for the White House to rename places and government departments in recent years — past renamings from the Gulf of Mexico to the Department of War and the Kennedy Center have all made headlines. But AI is different. A Pew Research survey from June this year found two-thirds of Americans interact with AI multiple times a week. Renaming this widely used technology directly collides with people's everyday language use.

Agencies Swap Terms On Paper, Not In Speech

Over the past month, most government websites have replaced the term quietly with no public fanfare. A new ICE tender requires contractors to build 'SI-powered' detention facility monitoring systems. The NIH website has posted a banner alerting users that both old and new terms may appear during the transition period. But many agencies are still lagging behind. The day after the executive order was signed, the Pentagon announced Project Meridian, a research initiative on future warfare, and still used the term AI in the accompanying press release. The FTC's press release last week wrote 'superintelligence' as one uncapitalized connected word, which does not meet the executive order's requirement of two separate words with both initial letters capitalized. HHS has not included the new term in most of its recent press releases. There are far too many AI-related offices across the federal government, and bodies such as the Chief Digital and Artificial Intelligence Office still keep their old names. The executive order also explicitly states that the new rule does not apply to regulations and documents released prior to the order's signing.

It is even harder for government employees to adjust their speaking habits. One researcher at a national lab interviewed by James O'Donnell slipped and said AI several times before awkwardly acknowledging his mistake. Defense Secretary Hegseth also used the term AI during an interview on October 3. On October 27, Trump called out the Wall Street Journal on Truth Social for refusing to use 'Super Intelligence', and hinted that new rules would soon ban the phrase 'artificial intelligence' in all government documents. Even Newsmax has not adopted the new term. The Associated Press already saw the consequences of noncompliance: after refusing to refer to the Gulf of Mexico as the 'Gulf of America', it was barred from covering some White House events. Word choice is starting to read as a public declaration of political alignment.

Giants Signed the Pledge Fast, But Are Unapologetic On Earnings Calls

On the same week the executive order was issued, the White House released a non-binding AI safety statement dubbed the White House Accord on Super Intelligence. The six signatories are Google's Sundar Pichai, Anthropic's Dario Amodei, Meta's Mark Zuckerberg, OpenAI's Greg Brockman, xAI's Elon Musk, and Nvidia's Jensen Huang. Three of them attended Trump's inauguration. Statements on paper are one thing, and unscripted earnings calls are another. A CNBC analysis of third-quarter earnings calls found that only Zuckerberg and Musk used the new term relatively consistently. Zuckerberg started using the word superintelligence publicly back when Meta rolled out its discounted AI offerings last year. Musk's companies use the term more than any other firm, and he even promised to rename SpaceX AI to SpaceX SI, but he still mixes SI and AI freely in his posts on X. Executives from Amazon, Microsoft and Alphabet all defaulted to the term AI on their latest earnings calls. Microsoft mentioned the 'Super Intelligence era' in its annual letter to shareholders. Huang has referenced the new term in prepared public remarks, but still slips back to AI in most casual settings. Salesforce's Marc Benioff argued that 'SIForce' sounds better than 'AIForce' — but that is, after all, just the name of a product for his own company. Federal government employees are now tasked with writing official renaming guidelines, and heads of all federal agencies must submit a formal legislative definition of Super Intelligence before the end of the month. Countless work hours and taxpayer dollars are being spent purely on swapping out a single set of words.

The Real Cost Is Public Conversation

O'Donnell is far more worried about the damage this move does to national public discourse around AI. He says he regularly overhears strangers talking about AI in grocery stores and coffee shops, most often discussing topics like AI-driven job displacement and AI-fueled future warfare. The conversations do not always sound optimistic, but they reflect that ordinary people want a say in how this technology reshapes their lives. Right now, public attitudes toward AI have not yet solidified along strict partisan lines. Once 'Super Intelligence' becomes a loyalty test, two people who are otherwise having a casual chat about whether chatbots benefit or harm mental health could hit an awkward impasse when one uses the term 'super intelligence' and the other still uses 'AI'. Midterm elections are just around the corner. Pew's survey found that for voters, AI ranks as a more important issue than environmental and climate policies, and Democratic voters even place it above abortion and gun policies. The national public conversation about AI has barely just started. Renaming the technology will not move that conversation forward — it will only add an unnecessary, partisan barrier to entry.

Source: MIT Technology Review, Author James O'Donnell, Original Link