White House 'Super Intelligence Force' adopts charter: a 120-day AI risk review and breach-reporting rules for labs

The White House superintelligence task force, which President Trump calls the "Super Intelligence Force," formally adopted its charter on Thursday, October 15, 2026, the day of its first official meeting, Semafor reported. According to Semafor, the adoption kicked off the 120-day window the force has been given to work out how the federal government should handle rapidly advancing AI. Semafor also reported that the force agreed to commission a report on frontier AI risk from the President's Council of Advisors on Science and Technology (PCAST), and that a senior Pentagon official will lead work on protocols for AI labs to report breaches and hacks.

This story is based on Semafor's October 15 exclusive, Trump's October 4 announcement as reported by CBS News and TechCrunch, October 4 and 5 reports from The Telegraph, Defense One and Nextgov, Axios's October 9 report on the force's incident mandate, and official White House and ODNI pages, checked October 16, 2026. The White House hasn't published the charter or a readout of the meeting.

The short version

• What happened: the force met for the first time on October 15 and formally adopted its charter, according to Semafor.
• The clock: a 120-day window, which Semafor says the charter's adoption kicked off. Reports from October 4 and 5 also described a 120-day deadline for a report on AI's risks and opportunities. No official start or end date has been published.
• The risk report: per Semafor, PCAST will assess AI risk in three areas: cybersecurity and critical infrastructure protection, biosecurity and dual-use applications, and model control and accountability.
• Breach reporting: Emil Michael, the Pentagon's under secretary for research and engineering, will develop protocols for labs to report breaches, hacks and other cyber issues, Semafor reports.
• Enforcement: unclear. A White House official told Semafor that export controls could be one tool.

Who runs the Super Intelligence Force?

Trump announced the force in a Truth Social post on Sunday, October 4, CBS News reported, after meeting leaders of major AI companies at the White House the previous week. The names and titles below come from official pages where possible, and from named outlets where not.

Last verified: October 16, 2026

No official list of titles has been published. CBS News's account of Trump's post uses neither "vice chair" nor "co-chair," and Axios's October 9 report called Ferguson, Michael and Kupor co-chairs.

Semafor lists other members, including Vice President JD Vance, Treasury Secretary Scott Bessent, Defense Secretary Pete Hegseth, OSTP Director Michael Kratsios, NSA Director Joshua Rudd and White House Chief of Staff Susie Wiles. The ODNI's leadership page lists Clayton as DNI, and Cairncross was confirmed as National Cyber Director in August 2025. Defense One called it "a unique move" for the sitting DNI, a former SEC chairman, to also lead domestic technology and economic policy discussions. Sen. Mark Warner told Defense One: "We still do not know what authority this new role will have."

What does the charter cover?

Semafor hasn't published the charter, and neither has the White House. Here's what's been reported about it:

• A 120-day window. On October 4, TechCrunch and The Telegraph reported a 120-day deadline for a report on AI's risks and opportunities, and Defense One reported the same on October 5. The Telegraph, citing the charter, added that the work could be extended beyond 120 days and its functions eventually handed to a government agency. Semafor wrote on October 15 that adopting the charter kicked off the 120-day window. No official start or end date has been published; counted from October 15, the window would end around February 12, 2027.
• Coordination, not new rules. The charter quoted by The Telegraph says the force "shall act as a co-ordinating body to ensure that the US remains at the forefront of SI development." TechCrunch reported it would "develop plans for responding to SI-enabled threats" while avoiding overregulation and regulatory capture.
• Breach notifications. Defense One and Nextgov reported that the charter includes streamlining breach notifications and strengthening the government's response capacity. The Telegraph said the force will review warnings about "breaches, hacks, jailbreaks and other security issues."

What do breach-reporting protocols mean for AI labs?

The force had already set expectations before its first meeting. On October 9, Axios reported that the force, responding to Anthropic's disclosures, told all AI companies to promptly disclose model incidents, fix any harm, and give "immediate and full transparency" to affected organizations and the public. Its statement said the process "is not optional" and that delays "will not be tolerated," but it named no penalties. The incidents behind it included a false homicide tip that a Claude model sent to Philadelphia police (our report).

Michael's protocols would turn that statement into a process. What's still unknown is the basic design: what counts as a reportable incident, how fast labs must report, and who receives the reports. Semafor reports it's unclear how compliance would be enforced or how current law would apply. The White House official's mention of export controls stands out, because Semafor notes the Commerce Department previously used them, temporarily, against Anthropic.

Other proposals go further. Microsoft President Brad Smith has called for mandatory incident reporting and an emergency brake (our story). On October 13, about 400 scientists asked for an international registry of containment failures (our story). The UK has said it is preparing responses to AI-agent incidents with its cyber agencies (our story).

Ferguson's FTC, meanwhile, sent civil investigative demands to OpenAI and Anthropic on October 14, Axios reported, and an FTC spokesperson confirmed the demands to CBS News (our story on the FTC demands).

What's confirmed, reported and uncertain

• Confirmed (official): Clayton is DNI (ODNI). Trump announced the force and named Clayton on Truth Social (quoted by CBS and TechCrunch). EO 14434 of September 29, 2026, replaced "AI" with "Super Intelligence" in executive-branch usage, but it doesn't create the force.
• Reported (Semafor exclusive, October 15): the first meeting and charter adoption, that the adoption started the 120-day window, the PCAST report and its three areas, the vice chairs' assignments, Cairncross's working-group role, the member list, and the enforcement question.
• Reported (TechCrunch and The Telegraph, October 4; Defense One and Nextgov, October 5): charter language, the 120-day report deadline and the breach-notification remit. TechCrunch cited the Wall Street Journal for Clayton as chair and the three vice chairs.
• Reported (Axios, October 9): the force's incident-disclosure mandate.
• Uncertain: the charter text, an official start and end date for the 120-day window, the PCAST report's deadline, the content and legal basis of the breach protocols, and whether they will apply to labs that didn't sign the accord.