Deyron Labs graphic: OpenAI partners with Hacktron to find vulnerabilities in OpenAI's systems, focusing on third-party service interactions, with extended access via Daybreak Red

What happened

On 6 November 2026 OpenAI published “Partnering with Hacktron to find vulnerabilities in OpenAI’s systems” in its Security category. The post announces a partnership with Hacktron, a company that builds AI agents for code review, penetration testing and red teaming. OpenAI engineers and Hacktron researchers choose together which areas to test first, with an emphasis on how OpenAI’s systems interact with third-party services. We read the post on openai.com on 7 November 2026.

Key details

  • Model access: OpenAI says it is extending Hacktron’s access to its frontier models through Daybreak Red. OpenAI’s separate GPT-6 Cyber announcement describes access through Daybreak Red as restricted to approved customers.
  • Background: OpenAI describes Hacktron as known through its Security Bug Bounty program.
  • Context from OpenAI: the post points to two announcements from earlier in the week, dated 2 November: that Cathedral researchers will work in OpenAI’s next security sprint, and that bug bounty rewards have been increased.
  • Quote: the post includes a statement from Zayne Zhang, CEO of Hacktron.
  • Not stated: the post gives no number of vulnerabilities found, no payment amounts, no duration and no terms of the agreement.

Why it matters

The announcement shows OpenAI pairing outside security researchers with its own frontier models, rather than relying only on internal teams and a public bug bounty. Its stated focus on interactions with third-party services points to the integration layer as an area of concern. For users and developers who connect other services to OpenAI products, the practical takeaway is limited for now: there are no findings to act on yet.

Caveats: this is OpenAI’s own description of the partnership. We could not verify any link between this partnership and earlier press reports about Hacktron and OpenAI systems, and the post does not mention them, so we do not draw one.

Disclosure: Claude, made by Anthropic, was one of the AI tools used to research and draft this article. Anthropic is a competitor of OpenAI. All claims are OpenAI’s unless stated otherwise.

Sources

  1. Partnering with Hacktron to find vulnerabilities in OpenAI's systems Primary source , OpenAI