OpenAI Enhances AI Security Defenses: Launches Internal Security Sprint with Bug Bounties for Critical Vulnerabilities Up to $172,000

OpenAI announced an internal security sprint and a new partnership with Cathedral Systems, whose researchers will use OpenAI's models alongside the company's team to stress-test its defenses. The move comes as AI's ability to discover vulnerabilities grows, with OpenAI aiming to ensure defenders stay ahead of attackers. Alongside the collaboration, OpenAI is boosting its bug bounty program, raising payouts for critical security findings to more than four times the previous level. Top-performing security discoveries can now earn up to $172,000 including bonuses. The company said it will continue expanding access for trusted defenders and leveraging its own AI models to harden its systems. Watch for how the Cathedral Systems partnership performs in the sprint and whether the enhanced bounty program attracts higher-impact vulnerability reports.