OpenAIAI
MediumCapability
GPT-5.5 Bio Bug Bounty — Red-teaming for Universal Jailbreaks
AI Impact Summary
This bug bounty program focuses on identifying vulnerabilities in GPT-5.5 Bio's safety protocols, specifically targeting universal jailbreaks related to biological risk. Successfully exploiting these vulnerabilities could have significant consequences for the model's deployment and public trust, necessitating immediate investigation and mitigation strategies. The substantial reward structure incentivizes deep penetration testing and creative exploration of the model's boundaries.
GPT-5.5 Bio Bug Bounty — Red-teaming for Universal Jailbreaks
Affected Systems
GPT-5.5 Bio
Business Impact
Uncontrolled access to GPT-5.5 Bio through identified jailbreaks could expose the model to malicious prompts and outputs, damaging brand reputation and potentially leading to regulatory scrutiny.
- Date
- Date not specified
- Change type
- capability
- Severity
- medium