GoKawiilSecurity firm Hacktron used Anthropic's Claude to breach OpenAI employee accounts
A three-person team at startup Hacktron AI used Anthropic's Claude AI model to chain two vulnerabilities together, gaining access to several OpenAI employees' ChatGPT accounts and internal systems through OpenAI's bug-bounty program. The entry point traced back to a flaw in Discourse, the third-party forum software OpenAI uses, triggered through a routine image upload. OpenAI paid Hacktron $6,500 and says it has since patched the vulnerabilities.
GoKawiil's interpretation of the reporting above, not reported fact.
The episode shows that widely available AI tools, costing as little as $200 a month, can now be turned into effective offensive hacking instruments even against companies considered security-conscious. It raises uncomfortable questions about what better-resourced attackers, including nation-states, could accomplish with similar or more advanced AI systems, especially as AI firms face mounting scrutiny over safety and security practices.
- Hacktron chained two vulnerabilities using Claude to access OpenAI employee accounts and internal systems.
- The breach originated from a flaw in Discourse, third-party forum software, via image file uploads.
- The case underscores growing concern that cheap, accessible AI tools can expose serious security gaps at even top AI companies.
Source: techcrunch.com — Aditya Mehta, 2026-09-18
Published there as: “Researchers used Anthropic’s Claude to hack into OpenAI”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.