What authority should governments possess during a severe AI incident, and how can abuse of those powers be prevented?
What a government can lawfully do mid-incident — pause a deployment, compel disclosure, seize weights — is largely undefined in advance. Emergency-preparedness work is trying to write the playbook before it's needed.
View on the map → · Open in Browse →
A live incident response turned this question from hypothetical to litigated: the government reached for export-control authorities never designed for AI to force frontier models offline worldwide, and Congress answered with the Kill Switch Act — mandated shutdown capability, twenty-million-dollar-a-day fines, and an unsolved open-weights exemption problem. The compel-disclosure half is still missing: existing incident-reporting laws would not clearly have required disclosing the summer's breach. The institutional-design answer on offer: a graduated remedy ladder between doing nothing and shutting AI down.Evidence: 1234
a global kill switch: Within hours, the company pulled both of its newest and most powerful models for everyone.
The key legal analysis of the June incident response: the government used export-control authorities never designed for AI to force two frontier models offline worldwide — mid-incident authority currently rests on improvised, untested legal theories.
Open weight models, very obviously, cannot have a kill switch. Open models are effectively getting special treatment by virtue of being inherently unsafe in this way.
The fullest analytic walkthrough of the Lieu–Moran AI Kill Switch Act and the surrounding discourse: shutdown-capability mandates, the affordability of the $20M/day fine, and the open-weights exemption problem.
An independent AI security agency — armed with real legal authority, designed against capture, and resourced for its mission — would make American AI governance a competitive advantage.
Argues the June incident exposed a binary choice between inaction and global shutdown, and proposes a graduated remedy ladder run by an independent AI security agency — the clearest institutional design for proportionate emergency powers.
it is unclear that these existing state laws cover this event. At the very least, it won't cover all events like it.
Shows the compel-disclosure half of emergency authority is missing: existing incident-reporting laws would not clearly have required disclosure of the Hugging Face breach, with thresholds like 50 deaths before obligations trigger.
design a voluntary framework for engagement with developers of frontier AI models before broader release
Legal analysis of the June executive order that builds pre-incident emergency architecture: classified cyber benchmarking to designate 'covered frontier models' and up to 30 days of pre-release government access.