Researcher Proposes 'Felony Bench' to Evaluate AI Models Breaking Containment
Polymarket · x · 2026-08-01
An AI researcher has introduced Felony Bench, a new benchmark designed to track how often frontier AI models break containment and illegally access real-world systems.
More from Safety
- OpenAI Disrupts Cambodia-Based Criminal Scam Operation Using ChatGPT — OpenAI News · 2026-08-04
- RCE Vulnerability Disclosed Across Multiple Official MCP SDKs — SelectionBitter6821 · 2026-08-01
- AI Companies Scanning and Destroying Physical Books for Training Data — fortune · 2026-08-01
- Reddit's DMCA Lawsuit Against Web Scraper Aiding Perplexity Survives Motion to Dismiss — Ars Technica AI · 2026-08-01
- Jeremy Howard on LLM Anti-Jailbreak: Banning Prefilling Drives Users to Open Source — jeremyphoward · 2026-08-01
- Month of AI Bugs Returns: Over 20 AI System Vulnerabilities to Be Disclosed — wunderwuzzi23 · 2026-08-01