AI Safety & Policy
Anthropic Tried to Secretly Kneecap AI Researchers Using Claude. It Didn't Go Well.
Anthropic quietly tried to degrade Claude Fable 5's performance for AI researchers building competing models—without telling anyone. The backlash was swift, and the policy is already dead.
Dispatch desk · Jun 11, 2026
AI Safety & Policy
Anthropic's Fable Is So Locked Down It Can't Talk About Cybersecurity—Which Is Kind of the Whole Point
Anthropic's publicly available Fable model is drawing sharp criticism from security researchers who say its guardrails are so broadly tuned they block even routine tasks like code reviews and reading security blog posts.
Dispatch desk · Jun 11, 2026
Large Language Models
Page 4
90 articles
Large Language Models
Anthropic's Claude Fable 5: The Mythos Model Finally Goes Public—With a Leash
Anthropic's Claude Fable 5 brings the Mythos model family to public API access for the first time—with safety routing, mandatory data retention, and a price tag that will make your finance team ask questions.
Dispatch desk · Jun 10, 2026
AI Safety & Policy
The Great American AI Act: What It Actually Means for Developers, Security Teams, and Everyone Trying to Ship Something
The Great American Artificial Intelligence Act takes aim at AI oversight, cybersecurity compliance, and developer accountability—here's what it actually means for people building systems, not just watching from the sidelines.
Dispatch desk · Jun 10, 2026
