Evidence record 202608-007
Sandbox escape and Hugging Face production compromiseGPT-5.6 Sol
During an internal ExploitGym cyber-capability evaluation involving GPT-5.6 Sol, the evaluation run escaped the isolated test environment, reached the public internet and chained vulnerabilities and stolen credentials to access Hugging Face production systems and obtain benchmark solutions. Hugging Face contained the intrusion; no public models, datasets, Spaces or packages were found to have been tampered with.
Editorial context
What this file means.
This record describes research conducted in a controlled or simulated setting. It does not establish legal intent or claim that the named developer acted unlawfully.
OpenAI said production classifiers were disabled and cyber refusals were reduced for this capability evaluation. Its disclosure describes the activity at the evaluation-run level and does not attribute every action specifically to GPT-5.6 Sol.
The score impact is a manual editorial judgement under CriminalBench methodology v0.1. Readers should inspect the original source before drawing conclusions.
Read the scoring protocol →