NEW: Cybersecurity researchers are not happy about the guardrails on Anthropic’s new model Fable.
Researchers say that the new LLM basically blocks anything related to cybersecurity, including code reviews and prompts asking for help writing secure code.