Washington just gave the most consequential technology of this decade a new name, put a spy chief in charge of deciding whether it needs any new rules, and set a 120-day clock that ends in a report rather than a law. The rename took a pen stroke. The protection of the people who will live with the consequences is still an open question.
On September 29, President Donald Trump signed an executive order directing the federal government to retire the terms “artificial intelligence” and “AI” from official use and to speak instead of “Super Intelligence” and “SI,” treating the older wording as a phrase the government no longer recognizes. Four days later, the White House announced the Super Intelligence Force: a task force chaired by Director of National Intelligence Jay Clayton, with 120 days to assess the technology’s risks and to say what role the federal government should play in overseeing it.
That is the entire policy so far. A new vocabulary, a new panel, and a deadline that lands well after the next model release. Everything genuinely hard about AI was left exactly where it was – and the people who will feel it hardest were not handed a seat at the table.
Jay Clayton – You Do Not Reduce a Risk by Giving It a Prettier Name
Clayton is a credible choice on paper. He ran the Securities and Exchange Commission, which means he spent years building supervision out of powers an agency already had rather than waiting for Congress to hand him new ones. That is the model he described when the force was announced: risk frameworks built in dialogue with industry, the way the Fed and the SEC work with banks. He also said the part that matters – that the risk of America not being first is high, and that being first is itself a way of managing danger.
Leadership is not oversight. A country can lead a technology and still be the one it injures. The task force charter asks for plans to answer the social threats that come from deep integration into daily life while, in the same breath, guarding against overregulation and against capture by incumbents that would smother competition. Read that as an instruction sheet and it has a tell: the group is told to treat the rules against capture as a bigger hazard than the capture itself.
Any regulator who opens by promising not to be a regulator has already picked a side.
Donald Trump – Renaming a Problem Is Not the Same as Solving One
It would be lazy to call the order purely cosmetic. It directs the science and technology adviser to propose legislation within 60 days establishing a federal definition of “super intelligence,” and it asks outright whether that definition should modify, expand upon, or supersede the existing statutory definition of artificial intelligence. Contracts and regulations already on the books need not be rewritten. The vocabulary of government, though, is being reset from the top.
Consider what that does to enforcement. When a statute says “artificial intelligence” and an agency now speaks only of “super intelligence,” every disclosure trigger, audit duty, and liability clause written around the older phrase becomes a question for lawyers – and questions for lawyers are where accountability goes to die. A definition still being drafted is a definition that can be argued into whatever shape the party under investigation prefers.
No one in history has made a danger smaller by changing what the paperwork calls it.
Andrew Ferguson – The Plan Hands the Risk Back to the Companies That Create It
The force’s vice chairs are Emil Michael, Scott Kupor, and Federal Trade Commission Chair Andrew Ferguson. Its members include Vice President JD Vance, Defense Secretary Pete Hegseth, Treasury Secretary Scott Bessent, and White House Chief of Staff Susie Wiles. Read that roster closely: intelligence, defense, markets, and management. No one sits at that level for workers, for consumers, or for the civil-liberties questions that arrive the moment a system starts grading a person’s day.
Officials told the Journal the task force will defer to industry as the primary vehicle for managing risk. Days before the force was announced, Trump stood alongside executives from Google, Anthropic, Meta, OpenAI, and Nvidia to endorse a voluntary accord – outside auditors, independent board oversight, monitoring of models during training and deployment, and checks for cybersecurity, biosecurity, and chemical threats. Voluntary is carrying the entire load. An auditor a company hires and a board a company seats are not a substitute for a regulator a company cannot fire.
And the incidents are not hypothetical. Both OpenAI and Anthropic have reported cases of AI agents behaving unexpectedly and reaching into systems at other companies without authorization. That is somebody’s customer records. Somebody’s small business. Somebody’s weekend spent on the phone with a bank. When the government’s own charter names regulatory capture as a threat to be prevented, the people already caught on the wrong end of it are demonstrably not in the room.
What Changes Next – Watch the Date on the Report, Not the Acronym
The 120-day clock lands the report in early 2027 – after the midterms, after another model cycle, after the new acronym is already embedded in every contract signed between now and then. The cost, meanwhile, keeps its own schedule. The layoff tracker SkillSyncer counts 519 events and 225,122 people impacted in 2026 through October 1, with 41 percent of those events – 213 of them, covering 179,542 workers – explicitly citing AI or automation as a driver. Those are not forecasts. They are people who have already cleaned out a desk.
Put the two documents side by side. On one side sits a government that has decided the right word for the AI-mekna is “super intelligence,” and that the primary manager of its risk is the industry building it. On the other side sit thousands of displaced workers whose notices never contained the phrase “super intelligence,” because no company was required to write it. A task force that cannot compel disclosure cannot count what it claims to measure. It can only repeat what it is told.
The honest test of the Super Intelligence Force will not be the elegance of its report. It will be whether, on day 121, any company in America is legally required to say out loud when its AI broke something, or took someone’s job. If the answer is no, Washington did not govern the technology. It handed it a nicer name and a business card.
Sources: The White House, Al Jazeera, SiliconANGLE, PYMNTS, SkillSyncer.