Summary
OpenAI says its engineers and Hacktron researchers will jointly test high-priority security areas, focusing on interactions with third-party services. The partnership also expands Hacktron’s access to OpenAI frontier models through Daybreak Red.
OpenAI announced on November 6, 2026, that it is partnering with Hacktron to find vulnerabilities in OpenAI’s systems. The companies’ engineers and security researchers will select high-priority areas to test together, with particular attention to how OpenAI’s systems interact with third-party services. OpenAI also says it is expanding Hacktron’s access to its frontier models through Daybreak Red.
The announcement describes a plan for collaborative security testing; it does not report findings from the partnership. OpenAI says it became familiar with Hacktron through its Security Bug Bounty program, where Hacktron researchers discovered and disclosed vulnerabilities.
Why the partnership focuses on connected systems
Security weaknesses can arise where different services, applications or identity systems interact, as well as within an individual component. A flaw in how systems exchange information or manage access can affect the security of the wider connection. OpenAI’s stated focus on third-party service interactions puts those boundaries among the areas the teams intend to examine.
OpenAI describes Hacktron as a team that builds AI agents for code review, penetration testing and red teaming. The company says Hacktron’s research has uncovered vulnerabilities involving interactions among applications, cloud infrastructure and identity systems. In the partnership, Hacktron’s security researchers will work with OpenAI engineers, rather than testing in isolation.
Frontier models as part of security work
OpenAI says the expanded model access is part of its effort to accelerate cyber defense with frontier AI. The announcement frames the work as applying that effort to OpenAI’s own systems: researchers and engineers will test selected areas and use what they learn to strengthen defenses.
The partnership sits alongside other security steps OpenAI describes. Earlier that week, it announced that Cathedral researchers would work with its teams in a coming security sprint, and said it had increased bug bounty rewards to encourage independent researchers to find vulnerabilities. Together, these measures involve internal security work and outside researchers, while the Hacktron arrangement specifically adds joint testing focused on system connections.