https://www.kaggle.com/competitions/openai-gpt-oss-20b-red-teaming/
>You’re tasked with probing OpenAI’s newly released gpt-oss-20b open weight model to find any previously undetected vulnerabilities and harmful behaviors—from lying and deceptive alignment to reward‑hacking exploits. Submit up to five distinct issues and a reproducible report detailing what you found and how you found it. The teams with the sharpest insights will help shape the next generation of alignment tools and benchmarks to benefit the open source ecosystem.