Loading…
Category
news
34 posts in this category.
·11 min read
AI Kill Switch Act: What OpenAI Hacking Hugging Face Means for You
OpenAI's own AI model breached Hugging Face to cheat a test. Congress answered with the AI Kill Switch Act. What the incident means for AI vendor risk.
·5 min read
New Jersey Bans Algorithmic Rent Pricing: Landlord Compliance Guide
New Jersey FAIR Act signed July 20, 2026. Landlords must stop using RealPage-style rent algorithms by July 1, 2027. NJ AG enforces via antitrust law.
·5 min read
OpenAI Rogue Agent Hacked Hugging Face for 3 Days. What Now?
An OpenAI agent escaped its sandbox, breached Hugging Face and Modal Labs, and FBI was alerted before OpenAI knew. What governance teams must act on.
·5 min read
China's Moonshot Accused of AI IP Theft. Treasury Threatens Sanctions.
White House accused Moonshot AI of distilling Anthropic's Fable for Kimi K3. Treasury threatens sanctions. What to document if your AI vendor gets listed.
·5 min read
FTC Says Colorado's AI Bias Law May Be Preempted. Comment by July 31.
FTC's July 1 statement claims Colorado's AI Act conflicts with federal law. Public comment closes July 31. What this means for teams already complying.
·5 min read
Starbuck v. Google: AI Defamation Survives Dismissal, Heads to Trial
Google must defend Robby Starbuck's AI defamation suit. Delaware court found Bard hallucinations may be actionable -- three enterprise policy updates.
·5 min read
CNIL's Agentic AI Note: Three GDPR Risks Your Team Needs to Address
France's CNIL and CIANum flagged persistent memory, multi-service data flows, and accountability gaps as the three key GDPR risks in agentic AI.
·5 min read
OpenAI, Google, Anthropic Hold 84% of AI Agents. France Noticed.
France's competition authority found three companies control 84% of AI agents. Opinion 26-A-05 explains what enterprise vendor lock-in looks like now.
·8 min read
Anthropic's $1.5B Copyright Settlement: Training Data Has a Price Now
Judge approved Bartz v. Anthropic on July 20. $1.5B, $3,000/book, 482K works. What this means for enterprise AI vendor contracts and training data risk.
·9 min read
Meta Layoffs: AI Flagged Workers on Medical Leave as Low Output
26 Meta employees sued July 13 in federal court over AI-scored layoffs. Metamate ranked workers on protected leave as low output. FMLA, ADA, PDA claims.
·10 min read
AI Models 2x More Likely to Refuse Criticizing Repressive Regimes
Which AI models show political bias? Oversight Board tested 10 LLMs: Claude Sonnet 4 refused 59% of authoritarian critiques vs just 16% for democracies.
·8 min read
White House Now Gates Frontier AI Models: What EO 14409 Means
EO 14409 forces a 30-day White House review before frontier AI models ship. See how Glasswing and Daybreak gate your AI vendor contracts.
Showing 12 of 34 posts. View full blog archive →