Trump’s new AI task force: who reviews risks, who can require answers

For someone whose private data are exposed, a government promise to study AI risks is different from a requirement that a company notify them. President Donald Trump’s new AI task force could shape future policy; its promised review does not itself create that kind of enforceable protection.

Trump announced on October 4 that Director of National Intelligence Jay Clayton would lead the new Super Intelligence Force. Reuters reported the announcement.

Its other announced leaders are Andrew Ferguson, chairman of the Federal Trade Commission; Emil Michael, the Pentagon’s chief technology officer; and Scott Kupor, director of the Office of Personnel Management. Reuters.

The government’s response involves three different jobs:

The panel will review reporting mechanisms for breaches, hacks and other AI incidents and recommend stronger federal responses under existing authority, according to Reuters’s account of Wall Street Journal reporting on its charter. Reuters.

Reviewing how incident reports reach the government does not, by itself, impose a new requirement that companies report every AI failure. Nor does a forthcoming review suspend obligations already in force.

Consider a health app that can combine information you enter with data from a fitness tracker. The FTC says such an app is likely to qualify as a personal-health-record vendor under its Health Breach Notification Rule. If a covered vendor suffers a qualifying breach of unsecured, identifiable health information, it must notify affected people, the FTC and, in some cases, the media.

That rule specifies who is covered and what triggers notice. It is an existing protection for particular health information, rather than a universal AI-incident reporting system. A task force can examine whether reporting arrangements work well; the duty to send a particular notice comes from the applicable rule. FTC guidance.

The FTC also has tools for obtaining answers from companies. Its civil investigative demands can require documents, testimony and written responses; recipients can challenge demands, and the agency can seek court enforcement. Section 5 of the FTC Act prohibits unfair or deceptive commercial practices. Enforcement uses administrative or judicial procedures, rather than following automatically from a policy recommendation. FTC authority overview.

A separate inquiry is already underway. On September 30, a senior FTC official told Reuters of an industry-wide investigation involving Anthropic, OpenAI and other AI labs, describing planned information demands and compelled testimony, including from the research group METR. An investigation gathers evidence; it does not establish wrongdoing. Reuters’s investigation report.

The announcement also follows a September 29 meeting with Trump at which AI companies agreed to establish voluntary standards, Reuters reported. Those commitments are another distinct element: a voluntary standard does not itself carry the force of a compulsory demand for evidence. Reuters.

The force’s name needs decoding, too. A September 29 executive order directs executive agencies, as far as the law permits, to replace AI terminology with “Super Intelligence” and “SI” in communications and other non-statutory documents. The order defines the renamed technology by reference to the existing statutory definition of AI, preserves agency authority and does not require changes to previously issued regulations. The new label records the administration’s vocabulary; it does not demonstrate that a system has acquired new capabilities.