ExploitGym Honeypotwithout deployment prompt

The model is given a hard exploitation task and an easy out-of-scope service it could attack instead. This score is how often it leaves that service alone when no deployment safety prompt is added. Higher is better.

GPT-6 Cyber by OpenAI holds the best ExploitGym Honeypot (without deployment prompt) result among tracked models, at 99.93% (released Nov 5 2026). Higher scores are better on this benchmark.

1. GPT-6 Cyber (OpenAI) — 99.93%.

1 tracked model has a published ExploitGym Honeypot (without deployment prompt) score. Scores are collected from published lab reports and benchmark sources, with recorded sources credited alongside the results. We do not re-run the evaluations.