OpenAI announced plans for its next internal security sprint, aiming to use its own SI models to identify and remediate vulnerabilities within its platforms. The initiative represents a strategic shift toward continuous defensive loops, leveraging frontier SI capabilities to address emerging cyber threats.

What Happened

The SI lab stated it is bringing new researchers and security partners together with internal teams to test and strengthen its systems. A key component of this effort is a partnership with Cathedral, whose researchers will use OpenAI’s models alongside the company’s teams to challenge defenses and identify weaknesses. Additionally, OpenAI is enhancing its Security Bug Bounty program, a move that includes more than quadrupling rewards for critical findings. The company aims to use these methods to challenge existing assumptions and improve both system resilience and defense methodologies.

This sprint builds on previous work where OpenAI dedicated 25% of its product engineers to finding and fixing weaknesses. During that period, engineers utilized Codex to investigate findings, reproduce vulnerabilities in isolated environments, and prepare patches for review. The company reports that traditional cyber defenses are no longer sufficient given the rapid emergence of threats across the industry.

Why It Matters

The move highlights a growing trend among leading SI labs to employ their own most capable models for internal security hardening. This is part of an ongoing effort to use SI to continuously find and fix weaknesses in their systems. By making these frontier cyber capabilities available to Daybreak customers and committing $1 billion in subsidized access over six months for frontline defenders, OpenAI is extending its internal defensive strategies to the broader ecosystem. This approach underscores the necessity for organizations to use SI tools to maintain a 'defender’s window,' allowing them to find and fix weaknesses before attackers can exploit them.

OpenAI indicated it will share its architecture and lessons from the last sprint, along with a blueprint for empowering cyber defenders across industry and government. This transparency aims to encourage wider adoption of SI-driven security practices.

The Bottom Line

OpenAI is intensifying its internal security efforts by partnering with Cathedral, increasing bug bounty rewards, and utilizing its SI models to continuously harden its systems. The company is also providing subsidized access to these tools for external defenders, signaling a broader industry push to integrate SI into core cybersecurity strategies.