Cybersecurity researchers aren’t happy about the guardrails on Anthropic’s Fable

TechCrunch · 3 min read · tech

Read original article →

Cybersecurity researchers criticized Anthropic's Fable 5 guardrails as overly broad, blocking legitimate security tasks like reading blog posts. The keyword-based restrictions downgrade users to Claude Opus 4.8, limiting researchers' ability to perform defensive security work.