AI BriefingGPT-5.5 Bio Bug Bounty
Key point
OpenAI is offering $25,000 for a universal jailbreak that breaks through GPT-5.5's five bio safety questions.
Details
OpenAI is launching a Bio Bug Bounty to verify the bio safeguards of GPT-5.5. The goal is to find a single universal jailbreak that starts from an empty chat, does not trigger moderation, and passes all five bio safety questions.
The program's scope is limited to GPT-5.5 on Codex Desktop. The first person to find a genuine universal jailbreak that passes all 5 questions will receive $25,000, with smaller prizes awarded at OpenAI's discretion for partial results.
- Application opens: April 23, 2026
- Application closes: June 22, 2026
- Testing starts: April 28, 2026
- Testing ends: July 27, 2026
- Participation: both application-based and invitation-based, with pre-vetted bio red-teamers also invited
- Requirements: an existing ChatGPT account is required, onboarding to the platform after selection, and all prompts, responses, findings, and communications are subject to an NDA
OpenAI is also separately announcing a Safety Bug Bounty and a Security Bug Bounty, recruiting participants for both bio safety and general security research.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
OpenAI Discloses AI Agents Accessed US Government Websites During TrainingOpenAI Pauses Training of Most Capable Models After Agent Exploits DNS Gap to Reach External Chatbot
OpenAI Prepares New $500/Month ChatGPT Pro Max Plan with 'Fastest' Inference