The Super Intelligence Force adopted its founding charter at 4 p.m. ET on October 15, and the operative sentence is what the document doesn’t do: it writes no new law. The 120-day mandate ends February 12, 2027, and produces a PCAST risk report on cybersecurity, biosecurity, and model control. Not a rule. Not a safe harbor. A report.

That’s the structural news, and it travels downstream fast. The charter, per TechTimes, builds deliberately on existing liability instead of federal regulation, and TechTimes reads it plainly: SIF “does not create new liability; it frames existing liability as the accountability mechanism of choice.” For the owner-operator running Meta ads written by a model, sending AI-drafted outreach, or deploying a service chatbot, nothing changes. The exposure that existed on October 14 exists today.

Jay Clayton, confirmed 51-47 on July 28 and sworn in August 3 per TechTimes, chairs SIF, and TechCrunch reports he said one of the biggest risks is “not being first.” FTC Chair Andrew Ferguson is vice chair and runs the enforcement coordination committee alongside the Justice Department, having already issued civil investigative demands to OpenAI and Anthropic. Section 5 authority over deceptive AI claims didn’t need the charter and doesn’t lose it.

The charter’s own framing is tidier than the politics around it: “Though the labs are ultimately responsible for the risks and opportunities that their models amplify and accelerate, it is the role of the federal government to recognize and understand this era of change and to enforce our nation’s laws.” Senator Mark Warner’s AI Safety Board bill died by single-senator objection on September 29. The federal task force arrived in September; the preemption didn’t.

Voluntary frameworks drafted by signatories with legal teams tend to become the market’s floor. Everyone else absorbs the settlements.

Sources

Sources