Semafor reported on Thursday that the White House AI task force adopted its charter after its first official meeting. That starts a 120-day window. The group agreed to ask the President's Council of Advisers on Science and Technology for an AI risk report. It covers three areas: cyber and infrastructure security, biosecurity, and who controls AI models and answers for them. Defense Undersecretary Emil Michael will lead an effort to develop protocols for labs to report breaches to the government. A working group will carry out the White House Accord on Super Intelligence, which major AI CEOs signed this month.
Semafor does not say what labs must report, and enforcement is not yet clear. It also says it is unclear how current law applies to companies that ignore the protocols. A White House official told Semafor that export controls could be one penalty. Commerce once briefly imposed them on Anthropic, after the company refused to give the government unfettered access to its models.
This is our reading, not Semafor's. The protocols cover what labs tell the government, not what they tell customers. A breach means an outsider got into a lab's systems. Customers cannot act on what they are not told. Buyers can ask vendors whether they will report incidents to customers, how fast, and whether the contract says so.