OpenAI has launched a Bio Bug Bounty program for GPT-5.5, inviting researchers to identify a universal jailbreak that can bypass its bio safety challenge. The program offers a $25,000 reward for the first successful jailbreak and runs from April 23 to June 22, 2026, with testing occurring from April 28 to July 27, 2026.
OpenAI is launching a Bio Bug Bounty program for GPT-5.5, specifically targeting researchers with expertise in AI red teaming, security, or biosecurity to identify universal jailbreaks that could bypass bio safety prompts. This initiative highlights a growing focus on ensuring AI safety in biology-related applications, offering a $25,000 reward for successful contributions. If you have relevant expertise, consider applying to leverage your skills in enhancing AI safety and security.