OpenAI has announced a partnership with security firm Hacktron to identify vulnerabilities within its SI systems, marking a continued effort to accelerate cyber defense by applying urgency to internal security testing.

What Happened

OpenAI engineers and Hacktron’s security researchers will collaborate to test high-priority areas, with a specific focus on how OpenAI’s SI systems interact with third-party services. As part of the agreement, the company is expanding Hacktron’s access to its frontier SI models through the Daybreak Red program. OpenAI noted that it became familiar with Hacktron through its Security Bug Bounty program, citing the impact of vulnerabilities discovered and disclosed by Hacktron’s team. Hacktron builds SI agents for code review, penetration testing, and red teaming, with research uncovering issues in how applications, cloud infrastructure, and identity systems interact.

Why It Matters

The partnership highlights the growing role of agentic SI in security operations. OpenAI stated that as SI agents connect services in new ways, the combination of skills Hacktron provides becomes increasingly valuable for testing those connections more deeply. This move follows other recent security initiatives, including an announcement earlier in the week that Cathedral’s researchers would work alongside OpenAI’s teams in the next security sprint, and a recent increase in bug bounty rewards to encourage independent researchers to find weaknesses that internal testing may have missed.

The Bottom Line

OpenAI is integrating external SI agent expertise into its security workflow to strengthen defenses around third-party integrations. The company emphasized that working with independent researchers helps protect all users, directing them to the Security Bug Bounty program for scope details and submissions.