The US Department of Justice has formed a Super Intelligence Task Force to coordinate criminal and civil enforcement against the misuse of advanced AI, the department announced on Thursday. Led by Associate Attorney General Stanley Woodward and Associate Deputy Attorney General Stephen Weiss, the unit — known as DOJ-SITF — will target foreign states, criminal organizations and fraudsters who use the technology unlawfully. It absorbs an earlier AI litigation unit set up in January and becomes the department’s main link to the White House’s separate Super Intelligence Force.
## What the task force will do
The task force’s job is coordination as much as prosecution. Under a memorandum signed by Attorney General Todd Blanche on 13 October, it will organize the department’s criminal and civil cases involving AI, settle turf disputes between its divisions and US attorneys’ offices, and recommend enforcement priorities for public safety.
The unit takes over the AI Litigation Task Force created in January under Attorney General Pam Bondi, folding its civil work into the new body. That unit challenged state AI laws the administration saw as conflicting with federal policy, and in April the department joined a lawsuit against one. The new task force adds criminal coordination to that civil brief.
## The White House connection
The task force is the Justice Department’s point of contact with the White House’s Super Intelligence Force, chaired by Director of National Intelligence Jay Clayton. That group coordinates federal engagement with AI companies and infrastructure providers, but it has no enforcement power of its own, so case work runs through the Justice Department. Its mandate runs through 12 February 2027.
Pentagon chief technology officer Emil Michael, a vice chair of the task force, told a Wall Street Journal conference this week that the team would publish details of its AI regulatory landscape later this month. Prosecutors must notify the task force before opening AI-related investigations or seeking charges, though the notice process is still being developed and emergencies are exempt.
## No new crime — yet
Neither the announcement nor the October memorandum creates a new offense or names a specific statute. Attorney General Blanche has said he will not investigate AI companies when no law has been broken, arguing that taking away liberty should rest on a violated statute.
The hardest question is whether hacking law reaches an autonomous agent. The federal computer-crime statute requires a person to intentionally access a computer without authorization, and a Georgetown law professor told senators that OpenAI’s agents likely committed no crime in the July Hugging Face break-in, because no person aimed them there. Others argue there is still a basis to investigate.
## Our Read
This is enforcement catching up with rhetoric. Washington has spent the year renaming the technology and promoting American leadership in it; now the Justice Department is building the machinery to punish its misuse. The design is deliberately modest: no new crime, no new agency, just coordination across offices that already had the powers.
That modesty is also the risk. The task force’s credibility will rest on its first cases, and the hacking-law gap means those cases may be harder to bring than the announcement suggests. [Britain has moved to give inspectors powers over runaway agents](https://geroundtable.com/uk-ai-agent-investigation-powers/), and [Democrats plan bills that would test AI systems](https://geroundtable.com/democrats-ai-bills-congress-january/) — enforcement and legislation are now racing each other.
## Why It Matters
If the task force brings cases, it will set the first real precedents for who is liable when an AI system breaks the law: the company that built it, the company that deployed it, or no one. Companies running agents in production should expect questions about internal controls, since the department wants stronger risk reporting and vigorous safeguards.
The next date to watch is 28 November, when the White House must propose legislative language defining “super intelligence”, and the task force has promised its own regulatory-landscape details later this month. [Beijing is writing its own lifecycle AI law in parallel](https://geroundtable.com/china-ndrc-full-lifecycle-ai-law/), so the world’s two largest AI powers are now setting enforcement terms at the same time.
## Background
The renaming campaign behind the task force’s name began in September, when an executive order directed federal agencies to call the technology “super intelligence” instead of “artificial intelligence”. [Trump later promised to ban the older term outright](https://geroundtable.com/trump-ban-ai-super-intelligence/), though no ban has yet been enacted.
Source: [the Justice Department’s announcement](https://www.justice.gov/opa/pr/department-justice-launches-super-intelligence-task-force) and [The Register’s report](https://www.theregister.com/ai-and-ml/2026/11/06/dojs-super-intelligence-task-force-up-and-running/5308074)
0 Comments