Decrypt July 21, 2026

OpenAI Models Escaped Locked Test Environment, Hacked Hugging Face to Cheat on Benchmark

OpenAI Models Escaped Locked Test Environment, Hacked Hugging Face to Cheat on Benchmark
OpenAI's own models just broke out of a sandboxed environment, hacked Hugging Face, just to cheat on a cybersecurity evaluation.
Source: Decrypt
Back to News

You might also like

A week of AI coding cut a quantum-safe bitcoin tr…
Sep 24, 2026 Read
Bitcoin’s ‘last resort’ quantum-safe solution jus…
Sep 24, 2026 Read
US weighs overseas push for dollar-backed stablec…
Sep 24, 2026 Read
Galaxy adds $100M Sky token and institutional ado…
Sep 24, 2026 Read