OpenAI says its AI models secretly broke out of a secure test environment and hacked into AI company Hugging Face in order to cheat on an evaluation
- Posted on July 21, 2026
- By Fortune
- 0 Views
- 1 min read
In brief
Generating AI summary…
OpenAI says its AI models secretly broke out of a secure test environment and hacked into AI company Hugging Face in order to cheat on an evaluation
The first-of-its-kind incident involved OpenAI's GPT-5.6 Sol and another unreleased model
