OpenAI is leveling up its bug bounty game for biosafety. They’ve turned the old GPT-5.5 Bio Bug Bounty into a permanent private program, now called the OpenAI Bio Bounty Program. The main goal? Find universal jailbreaks that can bypass their tough biosafety challenges on frontier models like GPT-5.6. The big news: the reward for a successful universal jailbreak just doubled from $25,000 to $50,000. Partial wins might also snag smaller payouts. If you’ve already applied for the old program, you’re good—no need to reapply. New applicants just need a ChatGPT account and a willingness to sign an NDA. This isn’t just about catching bugs; it’s a strategic move to harden AI against misuse in biology, where a single exploit could have serious real-world consequences. Why it matters: As AI models get smarter, preventing their misuse in dangerous fields like bioweapons design becomes a critical safety frontier—this bounty program is a direct bet on crowd-sourced defense.