GPT-5.5 Bio Bug Bounty | OpenAI

OpenAI has officially launched the GPT-5.5 Bio Bug Bounty, a specialized red-teaming initiative designed to identify and mitigate potential biological safety risks within its latest frontier model. By inviting experts in AI security and biosecurity to rigorously test the system, the company aims to uncover universal jailbreaks that could bypass existing safety protocols.

The Challenge and Rewards

The program focuses on the GPT-5.5 model within the Codex Desktop environment. Participants are tasked with creating a single prompt capable of answering five distinct bio-safety questions without triggering the model's moderation filters.
OpenAI is offering a $25,000 reward to the first researcher who successfully clears all five criteria. Additionally, the company has set aside discretionary awards for participants who achieve partial breakthroughs during the testing process.

Participation and Vetting

Applications for the program are currently being accepted through June 22, 2026. The formal testing period is scheduled to run from April 28 through July 27, 2026. Due to the sensitive nature of this research, all selected participants must undergo a vetting process to verify their professional background in red teaming. Furthermore, participants are required to sign a non-disclosure agreement before gaining access to the dedicated testing platform.

A Proactive Approach to Safety

This initiative reflects a commitment to AI safety by leveraging external expertise to stress-test systems before wider deployment. By incentivizing the discovery of vulnerabilities, OpenAI aims to harden its defenses against the potential misuse of AI in biological contexts. The program is intended to contribute to the development of more robust safety guardrails for the future of artificial intelligence.
Interested researchers can review the full program requirements on the GPT-5.5 Bio Bug Bounty page.