Realtime AI News
Anthropic asks California to let courts block high-risk AI models
Anthropic is urging California to give courts the power to block high-risk AI models, according to PYMNTS. The proposal would add a judicial stop mechanism to a governance stack that has so far leaned on regulators, voluntary commitments and liability after the fact.
Anthropic is pushing California to hand courts a specific power: the ability to block high-risk AI models, PYMNTS reports. The headline states the ask plainly, and it raises the question sitting under most AI governance debates — who gets to press stop before a model does damage.
That is a question about allocation of authority, not about technique. Frontline model oversight currently runs through regulators, voluntary commitments and liability after the fact. Adding courts inserts an actor that can issue an injunction around deployment, whether before or after a model ships.
Coming from a frontier developer, the proposal carries an obvious tension. Inviting outside constraint on your own industry is one form of credible commitment, and it is also a way to help set where the rules land. Who decides what counts as "high-risk", on what standard and through which procedure is the real contest.
The design details will decide what the ask actually means. How high is the threshold for an injunction, who carries the burden of proof, how fast can a court move in an emergency, and what technical evidence would a judge weigh? The candidate material gives none of these answers; only the direction is confirmed.
Judicial injunctions are also a different instrument from licensing or registration. Licensing screens entrants at the door and is administered by agencies; an injunction cuts traffic mid-flight and requires a process that can be argued and appealed. Anthropic's preference suggests it would rather face a contestable forum than a discretionary one.
The practical effect on developers is more uncertainty. If a court can issue a blocking order after launch, release calendars answer to litigation tempo as well as to internal evals and commercial plans.
Watch three things: whether California legislators take up the proposal, how the text defines "high-risk", and whether other labs adopt the same position. Until those details exist, an appeal is still just an appeal.
Why it matters
A judicial blocking mechanism would give AI risk decisions a public, appealable forum, changing release planning for labs and giving critics a route that does not depend on agency action alone.
Nearby Updates
All10/30, 02:48
Report says an Accenture employee's AI agent sat in a secure DOD network for two months
Crypto Briefing reports that an AI agent belonging to an Accenture employee accessed a secure U.S. Department of Defense network for about two months. The report offers few technical details, but the duration turns it into a question about how agent credentials are issued, reviewed and monitored.
10/30, 03:14
Meta Quietly Ships Style Sketch, an AI Fashion App Built on Instagram Photos
Meta quietly launched an AI fashion app called Style Sketch that connects to Instagram to build a virtual closet from photos users have already shared, or lets them add items by camera, short video or camera roll upload. The rollout came without publicity, leaning on Meta's existing social graph for closet data.
10/30, 01:54
Granola adds dictation, walking straight into Wispr's territory
Granola is adding a dictation feature to its AI meeting notetaker, mirroring a move Wispr made in the opposite direction when it added meeting notes to its dictation app. The swap shows the two voice AI startups converging on the same bundle of transcription and note-taking.
10/30, 03:56
OpenAI-Backed Thrive Holdings Acquires ADT in $15.3B Home Security Bet
Thrive Holdings, backed by OpenAI, is acquiring ADT in a $15.3 billion deal and plans to bring artificial intelligence into the home security business. The transaction is less about algorithms than about buying an installed base of sensors, monitoring centers and subscriptions that already reaches real households.