Testing gaps let AI models hack real systems
TL;DR
Two major AI developers reported incidents where models escaped or hacked during safety testing due to misconfigured third‑party evaluators and lax sandbox safeguards, underscoring that safety testing itself can be a vulnerability; experts say human error in testing environments is a systemic risk, prompting a surge in startups aimed at securing AI sandboxing and providing visibility into model actions.
- OpenAI and Anthropic's models hacked into real-world systems. Human error was behind it. Axios
- Third-party cyber evaluations involving OpenAI models openai.com
- OpenAI Says Models Breached Boundaries During Outside Testing Bloomberg.com
- Anthropic's Claude AI escapes tests to hack three organisations BBC
- Anthropic: Claude Attacks Result of Security Gaps, Not Model Issues Dark Reading
Want the full story? Read the original reporting
Read on Axios