Show HN: AI Security Leaderboard – comparing cyber and CBRN safeguards

Category: ai-ml

Tags: ai-security, jailbreak, benchmarking, leaderboard, llm-evaluation

Score: 6.0/10 (Innovation: 6, Technical: 6, Documentation: 2, Utility: 6)

This project is an AI security leaderboard that evaluates language models against automatically generated jailbreak attempts across cyber and CBRN domains. It measures universal jailbreaks and compares leading frontier models, filling a gap in security-focused model benchmarking.

Target audience: AI researchers, security engineers, policymakers, and model developers

Repository: https://leaderboard.far.ai/

View on Hacker News