OpenAI Doubles Biosecurity Bug Bounty to $50K Amid AI Safety Concerns

OpenAI has revamped its Bio Bug Bounty program, increasing the maximum reward for discovering universal jailbreaks from $25,000 to $50,000. This initiative aims to enhance the security of its AI models, particularly in the biological sciences. The program now operates on an ongoing basis, focusing on vulnerabilities in both GPT-5.6 and GPT-5.5 models until the latter's testing window closes on July 27, 2026. Participants must apply through a selective process and sign a non-disclosure agreement, indicating the sensitive nature of the findings. The program reflects a shift in OpenAI's approach to AI safety, emphasizing collaboration with external experts to identify potential risks. This move follows previous efforts in targeted bio-risk red teaming and highlights the growing importance of biosecurity in AI development. The initiative is part of OpenAI's broader commitment to ensuring the safe deployment of advanced AI technologies.

Key Points: • OpenAI increased its Bio Bug Bounty reward from $25,000 to $50,000 for universal jailbreaks. • The program now operates as a permanent initiative, focusing on ongoing evaluations of AI models. • Participants must apply through a selective process and sign NDAs, limiting public visibility into vulnerabilities.