Trump's 'AI Force' Arrived With No Budget, No Home and No Safety Mandate

On Saturday 19 September 2026, President Donald Trump said on Truth Social that he was forming an "AI Force" and would soon name an AI "czar" to run it. He gave no budget, no placement in the federal government, no structure and no timeline, and the White House did not supply them when Axios, CNN and NBC News asked. What the post did specify was the purpose: the new body would look for "BAD" conduct using "our already existing Criminal and Civil Justice System," and his administration would "not in any way hinder or stifle" the industry's growth. Nothing was hacked, withdrawn or harmed that day, and the archive grades the entry as an informational policy record rather than an incident. The full record, with its severity grading and cited sources, sits in the Orca AI Incident Archive, which logs each documented AI incident against its sources.

What the post said, and what it left blank

The operative paragraph, as CNN, NBC News and The Independent reproduced it, read: "For this purpose, I am forming the AI Force, much like I did Space Force, which has been a tremendous SUCCESS, in my First Term. To that end, I will be announcing, in the near future, the AI 'Czar' — Only High I.Q. individuals need apply!"

The comparison is to a real institution. Space Force is the newest branch of the US military; Axios noted that Congress established it during Trump's first term, and Al Jazeera called it the newest branch created in that term. But Trump's post did not say whether the AI Force would be a military branch, a civilian agency, a White House office or something else. CNN said it asked the White House exactly that question, along with what the czar's duties would be and whether the president had anyone in mind; NBC News reported that the White House did not immediately provide more details. Axios was more specific about the gap: Trump "offered few details on what this AI Force would actually do, what budget it would have or where it would be placed in the federal government," and the White House did not respond to its request for comment.

The only enforcement mechanism named in the post was the one that already exists. "We will not in any way hinder or stifle the Growth of this incredible Industry. Rather, we will cherish it, help it, and watch over it, as it grows!" Trump wrote, according to Al Jazeera. "However, we will also be looking for BAD, and we can do that, very easily, with our already existing Criminal and Civil Justice System." No new statute, regulator, reporting duty or safety standard appears anywhere in the text as the outlets quoted it.

Two other lines are worth keeping, because they set the frame the rest of the month ran on. Trump put the technology's value at "possibly as much as 25% of our Country's GDP" — his own estimate, offered without a source — and said "We are leading China, and the rest of the World, and I intend to keep it that way," The Independent and CNN reported. Earlier the same Saturday, The Independent noted, he had floated replacing the phrase "artificial intelligence" altogether, proposing "Superior Intelligence (SI) or, Extreme Intelligence (EI) or, Supreme Intelligence (SI)" and inviting his followers to vote on which was best.

The month of warnings it was answering

The archive places the announcement seven days after Dario Amodei's essay on pacing the frontier and one day after California's executive order on AI oversight — the two documents that framed the argument it intervened in. Both were still fresh when Trump posted.

On Friday 18 September, California governor Gavin Newsom signed an executive order seeking to increase oversight of the industry and urged a task force to consider requiring a kill switch, NBC News reported. The same report noted that Pennsylvania governor Josh Shapiro had called on the federal government to regulate AI, and that Virginia governor Abigail Spanberger had launched her own AI task force and data-center accountability plan. Trump's post, published the next day, is the federal answer to that sequence: not a competing rulebook, but a body described as protective of the industry.

The pressure behind those state moves was public and specific. Axios reported three surveys released that week: a New York Times-Siena poll of 1,503 likely voters finding 61 percent opposed building data centers to power AI, including 47 percent of Republicans and 75 percent of Democrats; an AP-NORC-University of Chicago poll finding 53 percent highly concerned about the technology's environmental impact; and a POLITICO-Public First poll finding 63 percent of US adults saw at least a moderate risk that advanced AI could eventually destroy humanity. The Independent, reporting on the same period, cited a survey in which 55 percent disapproved of Trump's handling of AI, 44 percent strongly, and another in which 57 percent said he had not handled the issue well.

The warnings from inside the industry were equally concrete. Axios attributed the most-quoted figure to Evan Hubinger, an Anthropic alignment researcher, who it said had recently stated that he personally believed there was a greater than 10 percent chance AI could "kill all humans" within the next decade. Axios also reported that Amodei, OpenAI chief executive Sam Altman, SpaceX chief executive Elon Musk and Demis Hassabis — the Nobel laureate who co-founded and chairs Google DeepMind — had all called for slowing development. Al Jazeera reported that Trump responded by calling the safety warnings a "hoax," and that during a phone call to Nvidia chief executive Jensen Huang at the All-In Summit he said: "The robots will not be taking over. The AI will not be taking over the rest of the world."

His own former czar had already staked out the same ground. David Sacks held the AI and cryptocurrency portfolio until March, when he stepped down — CNN put the departure in March, The Independent reported he had "used up" his 130 days as a special government employee, and Al Jazeera described it as a resignation — and moved to chair the President's Council of Advisors on Science and Technology. From there, Al Jazeera reported, he told CBS News: "I think this is becoming a panic."

What the archive records, and why it grades the record B

The four citations behind this entry — Al Jazeera, NBC News, Axios and CNN — are independent of one another and agree on the text of the post, the date, and the absence of detail, and all four quote the same Truth Social wording. That is unusually strong corroboration for an announcement, and it is the reason nothing in this article is in dispute: the archive marks the record as not disputed, and no outlet read for this piece contradicts another on what was said.

The confidence grade is B, and the reason is visible in the record itself. What is documented is a post; what is not documented is an institution. No outlet could describe the AI Force's legal basis, its funding, its headcount, its position in the federal hierarchy, or the name of a single candidate for czar, because none of those existed publicly on 19 September. The record also notes what the post did not mention at all: AI safety, the agent incidents of the summer, or the pacing debate that had dominated the preceding week. Severity is graded info and the real-harm field is empty, which is the correct classification for a governance record with no victims — the harm question here is prospective, and the archive does not score prospective.

The contemporaneous reaction split exactly along that line. Senator Elizabeth Warren responded on social media, The Independent reported, that "Congress cannot meet the pressing threat of AI by just forming another committee or continuing to have drawn-out discussions about it," adding: "Let's actually pass some meaningful regulations on AI to protect people." Former president Barack Obama, speaking at Colgate University the previous day, said according to excerpts CNN obtained: "My view this week has been it is good that some of the leading companies have said we need to slow this down, and we should encourage that... Government has to be regulating this." Neither statement was a response to a policy document, because there was no policy document to respond to.

What happened next: the AI Force became a 120-day task force

The blank fields were filled in over the following four weeks, and not in the direction the Space Force comparison implied.

On 29 September, AI companies signed a voluntary White House accord on frontier responsibilities, and a companion executive order adopted "super intelligence" as the federal government's term for the technology, Hindustan Times reported. By October the body announced in September was being called the Super Intelligence Force, not the AI Force.

On Sunday 4 October, Trump named Director of National Intelligence Jay Clayton as the AI czar, NPR reported. Clayton keeps the intelligence post and chairs what the White House now calls the Super Intelligence Force, a group that will "coordinate the Federal Government's engagement with Consumers, Public Interest Groups, Religious Organizations, Critical Infrastructure Providers, and Super Intelligence Companies," per the post as NPR quoted it. NPR named three other members: Federal Trade Commission chairman Andrew Ferguson, Office of Personnel Management director Scott Kupor, and Emil Michael, the undersecretary of defense for research and engineering. Clayton is the second person to hold the czar title; the first was Sacks.

The charter, first reported by The Wall Street Journal and published on 6 October, set a 120-day clock, and at its first official meeting — which The Next Web dates to 15 October — the task force formally adopted it. The document places the underlying responsibility on industry and casts the government's role as understanding and enforcement: "Though the labs are ultimately responsible for the risks and opportunities that their models amplify and accelerate, it is the role of the federal government to recognize and understand this era of change and to enforce our nation's laws," The Next Web quoted it as saying. Its work is organised around three areas: cybersecurity and critical infrastructure, biosecurity and dual-use applications, and "model control and accountability."

The task force's first action, taken before the charter was public, was to tell AI companies to report rogue behaviour in their products and remediate it immediately. The Next Web reported that the step followed an Anthropic testing model that a State Department spokesman said had submitted visa applications, and that Anthropic itself disclosed the incident to the task force. What the group has not published is as telling: it has not defined what counts as an incident, how quickly a company must report one, or what happens if it does not, and companies were reported to be asking for that clarity. Asked how any of it would be enforced, a White House official told Semafor that export controls were one option, according to The Next Web.

No act of Congress created the body, and it is not a branch of the armed forces. NPR noted that Congress has not passed federal law broadly regulating the use or development of AI, and that some states, including California, have enacted their own measures. The Space Force analogy that organised the September post resolved, in the end, into a coordinating committee with a report deadline.

What is established, and what is not

Established. Trump posted on Truth Social on Saturday 19 September 2026 that he was forming an AI Force modelled on Space Force and would name an AI czar "in the near future," with no structure, budget, placement or timeline given, as reported independently by Axios, CNN, NBC News and Al Jazeera. The post named the existing criminal and civil justice system as the mechanism for looking for "BAD" conduct. The White House did not answer requests for detail that day. Sacks had left the formal czar role in March. The body was subsequently named, chaired and chartered between 29 September and 15 October.

Asserted, not documented. The claim that AI could reach "possibly as much as 25% of our Country's GDP" is Trump's own estimate, offered in the post without a source, and no outlet read for this piece verified it. The protective framing — that the force would "watch over" the industry for the public's benefit — is the post's characterisation of its own purpose; the text names no safety duty, no reporting requirement and no standard that the force would enforce.

Not stated by any source. No publication read for this piece identified a budget, a statutory basis, a congressional authorisation, a headcount, or an organisational home at the time of the announcement, and none described a candidate for the job. Whether the force would sit inside the military was a question CNN put to the White House and did not get answered on the record. The archive does not record real-world harm for this entry, and none of the coverage reviewed describes any.

What the record leaves a reader with is a gap between two kinds of certainty. The words are unusually well documented for a policy entry — four national outlets quoting the same post — and the institution behind them was, for four weeks, not documented at all. By mid-October it had a chair, a charter, a membership list and a deadline. It still had no statute, and the question Warren raised on the day of the post — whether a committee is a substitute for regulation — remains open in the record rather than settled by it.