DOJ's Super Intelligence Task Force absorbs the unit created to fight state AI laws
The Justice Department announced DOJ-SITF on November 5. The two-page memo behind it merges the January AI task force, tells prosecutors to give notice on request, and never defines Super Intelligence.
On November 5, the US Justice Department announced the Super Intelligence Task Force, or DOJ-SITF, framing it as a law-enforcement effort against foreign states, transnational criminals and fraudsters who misuse AI. Most coverage stopped at the name, which uses President Trump’s preferred replacement for “artificial intelligence.”
The more consequential text is the two-page memorandum signed by Attorney General Todd Blanche on October 13, more than three weeks before the announcement. It gives one body both criminal enforcement over AI misuse and the department’s existing campaign against state AI regulation, and it adds a stated mission to promote what it calls American SI dominance.
What the memo actually creates
The memo is addressed to all component heads and US Attorneys, and it cites Executive Order 14434, “Inaugurating the Era of Super Intelligence,” of September 29, which directed the creation of the White House Super Intelligence Force. DOJ-SITF is the department’s principal point of contact for that body, whose chair is Jay Clayton.
Its six listed functions are organizing criminal and civil enforcement involving SI; deconflicting cases among components and US Attorneys’ Offices; recommending enforcement priorities; acting as the liaison to the White House force; incorporating the earlier AI task force and coordinating affirmative civil litigation on SI; and developing prosecution and compliance guidance while advising on legislative or regulatory proposals.
The Attorney General or two designees, Associate Attorney General Stanley Woodward and Associate Deputy Attorney General Stephen Weiss, serve as co-chairs. Members come from the Antitrust, Civil, Criminal and National Security divisions, the National Fraud Enforcement Division, the FBI, the Office of the Solicitor General and the Attorney General’s Advisory Committee. The presence of Antitrust stands out, as nothing in the public mandate mentions competition, and the memo does not explain the seat.
The merger most coverage missed
A footnote says the Artificial Intelligence Task Force “announced on January 9, 2026” ceases to exist as a separate body, with its matters, work product and personnel moving to DOJ-SITF. That date matches the January 9 memo from then-Attorney General Pam Bondi creating the AI Litigation Task Force, which Executive Order 14365 charged with the sole job of challenging state AI laws that conflict with federal policy, on grounds such as preemption and the Commerce Clause.
The new memo does not repeat that purpose or mention states at all. It does retain “affirmative civil litigation with respect to SI” and tasks the group with advising on proposals “that promote SI innovation without burdensome regulation.” The press release describes only the threat side of the mandate. Read together, the same office now handles AI fraud cases and any federal suits against state AI statutes, and it has a written instruction to favor light regulation when advising on policy. Whether that changes how aggressively the department sues states is not yet knowable, but companies and state attorneys general should not treat this as a purely criminal unit.
Notice requirements are narrower than reported
The notice language is softer than a clearance requirement. Components and US Attorneys’ Offices “should, from time to time at the request of the DOJ-SITF,” notify it when seeking to open an investigation and before seeking charges or finalizing a resolution in matters involving SI. A footnote says the notice process should not delay action and is not necessary in exigent circumstances, including imminent public-safety threats or risk of losing evidence. It also says the memo creates no enforceable rights.
Even so, this is a centralizing tool. A small group in main Justice can ask to see AI-related cases before charges or settlements, which gives it visibility into, and potential influence over, prosecutions that US Attorneys’ Offices have traditionally run on their own. How often “from time to time” turns out to mean will determine how much it matters. The memo leaves the process to be developed.
An undefined term carrying legal weight
Neither the memo nor the press release defines Super Intelligence. MLex describes it as the president’s preferred term for artificial intelligence, and Engadget notes the president’s stated plan to ban the older phrase. In this document it appears to mean AI generally rather than any capability threshold. The practical result is that “matters involving SI” has no boundary: a phishing case using a chatbot, a deepfake fraud, a model theft and a dispute over a state safety statute could all qualify. For comparison, state-level and proposed AI safety laws tend to define covered systems by compute, revenue or harm thresholds. This one leaves the scope to the department’s discretion.
What companies should take from it
The one forward-looking signal for industry is compliance. Weiss said in the release that the department wants to ensure companies “exercise vigorous internal controls so that SI technology behaves as intended,” and the memo calls for guidance on evaluating compliance measures. DOJ’s corporate enforcement guidance has historically shaped what companies build, because charging decisions weigh the quality of a compliance program. If that pattern holds, AI-specific expectations (monitoring, misuse response, access controls, and documentation of model behavior) could arrive through prosecutorial guidance and not through legislation.
Three things to watch: the first published guidance on prosecuting SI-related conduct and evaluating compliance; whether DOJ files or joins suits against state AI laws under the new banner; and the first case publicly attributed to the task force. Politics may also intervene. Gizmodo and Engadget both expect congressional Democrats, after their midterm gains, to open tech investigations when the new Congress sits in January, which would put a department that has pledged to promote AI dominance in an awkward position if asked to police the same companies.
Get Insanely Good at AI
The book for developers who want to understand how AI actually works. LLMs, prompt engineering, RAG, AI agents, and production systems.
Keep Reading
gr.Workflow Turns AI Pipelines Into Deployable APIs
Learn how to model, debug, expose, and deploy multi-step AI pipelines with Gradio Workflow and daggr.
The New York Times and three Tribune-group papers sue Brave, targeting the search API that feeds AI companies
Two November 2 complaints in Manhattan federal court accuse Brave of selling news through its Search API and Ask Brave, bypassing paywalls and replacing publisher ads. The tests cited show near-complete articles returned through an endpoint built for LLMs.
The FRONTIER Act Already Answers OpenAI's Antitrust Question
OpenAI publicly backed the outside-evaluator provision of the FRONTIER Act, H.R. 9925, the July bill that also contains a limited antitrust exemption for good-faith AI safety coordination, the exact legal cover it asked Congress for.
Sanders and Casar Introduce a Bill to Ban Superintelligence, With 20-Year Prison Terms
Senators Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act on September 3, which would permanently prohibit AI matching or exceeding human cognitive performance and impose criminal penalties of up to 20 years.
G20 Unanimously Endorses the US Light-Touch Carolina Principles on AI
All G20 members, including China, endorsed the non-binding Carolina Principles for AI governance at the Chapel Hill ministerial on September 2, favoring sector-specific rules over comprehensive regulation.