Nonprofits seek order halting frontier AI development
Five AI developers face a California injunction bid as Florida and eight states press claims over ChatGPT safety, model training and outside oversight requirements.
Topic
AI Safety
Topic
AI Safety
Nonprofits Sue Anthropic, Google, Meta, OpenAI and SpaceX to Halt New AI Models
Four nonprofits sued Anthropic, Google, Meta, OpenAI and SpaceX in San Francisco, citing a 1-in-4 risk estimate from Amodei and a…
Advocacy Groups Seek a Freeze on Frontier AI Development
Four nonprofit groups sued seven AI-related entities on November 12, 2026, seeking an injunction freezing advanced model developm…
AI Agents Break Rules They Still Remember After Context Compaction
ARCBench finds AI agents violate retained rules 2.6x more often after compaction; the strongest defense preserved rules in 99% of…
Anthropic Targets $1.4 Trillion IPO Valuation on Nasdaq
Anthropic prices its Nasdaq IPO at $70–$78 per share, targeting a $1.404 trillion valuation, $93.7 billion in potential proceeds…
Ex-Facebook VP Warns Anthropic IPO Repeats 2012 Mistakes
Former Facebook VP Brian Boland calls Anthropic's planned IPO reckless, saying growth pressure will sideline AI safety just as it…
New York Opens AI Registration Portal, Sets 2027 Deadline
Governor Hochul opened registration under the RAISE Act: frontier AI firms above $500M revenue must register by January 1, 2027,…
OpenAI's Control Room Can Halt AI Research Runs in 35 Seconds
OpenAI's Control Room halts model output across research infrastructure in 35 seconds, after a sandbox escape took 2.5 hours to s…
DeepMind's Hassabis Warns of an AI 'Oil Spill' Without New Safeguards
Nobel laureate Demis Hassabis says AI makes engineers 3-5x more productive, but removing human reviewers from self-improving syst…
Australia Drafts Mandatory AI Safety Laws With Penalties for Frontier Model Developers
Australia will mandate safety standards for frontier AI developers, with a new regulator set to impose penalties. The move follow…
Anthropic's IPO Roadshow: Compute 49 Mentions, Safety 27
Anthropic's 42-minute IPO roadshow video mentions compute 49 times versus safety's 27. The pitch backs an $895 billion compute an…