OpenAI has hired an outside security firm to attack its own systems. Hacktron, a company that builds AI agents for code review, penetration testing and red teaming, will work with OpenAI engineers to find holes before real attackers do, and will get broader access to OpenAI’s frontier models for the job.
The two teams will pick the riskiest areas to test together. The stated focus is the points where OpenAI’s products connect to third-party services, the kind of link an AI agent uses when it logs into a calendar, a code host or a cloud account for you. Security teams worry about those joins because they are where a flaw in one company’s system can become a way into another’s.
Hacktron’s access comes through Daybreak Red, OpenAI’s programme that puts its strongest cyber-capable models in the hands of security researchers. OpenAI says it came across the firm through its Security Bug Bounty programme, where Hacktron researchers had reported vulnerabilities, and says it was impressed by their impact.
Hacktron’s own announcement frames the deal as a bet on pairing people with AI tools. Chief executive Zayne Zhang wrote that the future of cybersecurity “won’t be defined by who has the largest security team, but by who can best combine human creativity with AI-driven capabilities.”
OpenAI’s post on the partnership was published on 6 November. The announcement does not say what the testing found, and neither company has said how long the work will run. Independent researchers can still report problems through the bug bounty programme.
