Broad Power, No Regulators: The Leaked Charter of Trump's 'Super Intelligence Force'

The White House’s new “Super Intelligence Force” — the task force President Donald Trump announced on October 4 to steer the federal government’s response to artificial intelligence — has been granted sweeping authority to decide how both the tech industry and Washington should handle threats to national security and civil liberties posed by fast-evolving AI, according to a charter obtained by POLITICO on October 6.

Three people familiar with the task force’s internal discussions confirmed the document’s authenticity to the outlet, speaking on anonymity to share details not yet public. The Wall Street Journal had first reported parts of the charter a day earlier, when it revealed that the group faces a hard 120-day deadline to deliver its first report on AI risks and opportunities to the president.

The leak fills in the most consequential missing piece of an announcement that, until now, had been long on branding and short on substance.

What the charter actually says

According to the document, the task force is charged with identifying “measures to improve the preparedness and response capability of both the government and the private sector to threats to the homeland and the American people.” That mandate covers AI-enabled cyberattacks, the use of AI in critical infrastructure attacks, and the surveillance risks that advanced models pose to ordinary Americans.

The panel’s report is directed at four specific areas: potential threats to civil liberties, cybersecurity, standards for advanced models, and — in a line that signals the administration’s deregulatory instincts — the impacts of possible overregulation.

The membership tells its own story. Director of National Intelligence Jay Clayton, confirmed as spy chief in August and now doubling as the administration’s “AI czar,” chairs the group. He is joined by FTC Chairman Andrew Ferguson, Under Secretary of Defense for Research and Engineering Emil Michael, and Office of Personnel Management Director Scott Kupor. All four report directly to the president and his chief of staff.

Each member also owns a liaison portfolio. Michael, the Pentagon’s research chief, coordinates with AI companies; Kupor, the government’s personnel chief, handles critical infrastructure operators; and Ferguson, the FTC chair, is the group’s channel to consumer groups and advocacy organizations.

The gaps are the story

What the charter does not include may matter more than what it does. Not a single seat at the table is held by a dedicated AI regulator, and the government’s own cybersecurity agency is conspicuously absent: the Cybersecurity and Infrastructure Security Agency (CISA), the body nominally responsible for defending federal networks and critical infrastructure, “is nowhere to be seen” in the project, as Cybersecurity Dive put it. It remains unclear what role, if any, the government’s cybersecurity experts will play.

The four-person panel also lacks any member with deep, career-length AI safety research credentials. Clayton is a former SEC chairman and U.S. attorney; Ferguson is an antitrust lawyer; Michael is a venture capitalist and former Uber executive; Kupor is a former Andreessen Horowitz partner. The composition reads less like a technical safety board and more like a coordination and outreach body — which, given the charter’s own language about “overregulation,” appears to be by design.

It is also a break from precedent in another sense: the AI czar role effectively replaces David Sacks, Trump’s first “AI and crypto czar,” who served only part-time before stepping down in March after hitting his allotted 130 days.

Self-regulation over statute

The task force is the administration’s latest and most visible attempt to head off AI’s security threats — from faster, more sophisticated hacking to custom bioweapons — without actually regulating the AI industry. During a September 29 meeting with executives from major AI companies, Trump promised he would “never stifle the growth of technology that will be bigger than the industrial revolution.”

The pattern is consistent. In June, Trump signed an executive order creating a pre-release government review process for frontier AI models — but made participation voluntary. At the September meeting, executives pledged to improve internal testing procedures and submit their models to third-party reviews. The new charter follows the same logic: convene, coordinate, report, but do not mandate.

Internal divisions remain. National Cyber Director Sean Cairncross and Treasury Secretary Scott Bessent have reportedly clashed over how aggressively the government should oversee the industry, and Senator Mark Warner of Virginia, the ranking member on the Senate Intelligence Committee, said in a statement that the administration had “resisted meaningful safeguards without offering a serious alternative for addressing the national security risks posed by increasingly powerful AI systems.”

Why it matters

The 120-day clock started this week. That means the first authoritative statement of how the U.S. government intends to define, measure, and respond to AI risk — under an administration that has already renamed artificial intelligence “super intelligence” in official documents and dismissed superintelligence fears as a “hoax” — is now being drafted by four officials, none of whom run an AI safety program, with broad authority but no regulatory teeth.

For AI labs, the liaison structure offers a predictable, lightweight interface with Washington: one Pentagon channel for frontier developers, one FTC channel for consumer advocates. For critics, the charter confirms their fears that the response to AI risk will be voluntary, industry-led, and advisory — a report due in February 2027 that may recommend little more than the status quo.

Either way, the document now exists in public. How its four authors use the next four months will say a great deal about whether “Super Intelligence Force” becomes a genuine coordination body or just the latest rebrand in a year full of them.