Security researchers used Anthropic's Claude to hack OpenAI's internal systems in under 72 hours
The breach occurred through OpenAI's community forum, exploiting chained vulnerabilities. The exploit was completed within 72 hours of Opus 5's release. Hacktron AI's team participated in OpenAI's bug-bounty program.

Security researchers used Anthropic's Claude to hack OpenAI's internal systems in under 72 hours, revealing significant vulnerabilities in the company's defenses. The breach was carried out by a three-person security team at Hacktron AI as part of OpenAI's bug-bounty program. The attack exposed weaknesses in OpenAI's community forum, where chained vulnerabilities allowed access to employee accounts and code repositories.
The exploit was completed in less than 72 hours, a timeframe that highlights the rapid evolution of AI-driven cyberattacks. The attack became possible only after the release of Anthropic's Claude Opus 5, which provided the necessary capabilities to execute the breach. OpenAI confirmed that the issues identified by the researchers have since been resolved.
The use of Anthropic's Claude Opus 5 in the attack underscores the growing role of advanced AI models in enabling sophisticated cyber threats. The breach took place over a short period, demonstrating how AI can significantly reduce the time and expertise required for such exploits. This incident has raised concerns about the security of AI systems and the potential for similar attacks on other major AI firms.
The consequences of this breach include increased scrutiny on AI security practices and potential regulatory responses. Companies may face higher costs to strengthen their defenses, and there could be a shift toward more stringent governance frameworks. Market reactions may also influence investment in AI security solutions, as stakeholders reassess risks associated with AI-driven vulnerabilities.
While the situation remains under development, the incident has sparked discussions about the balance between innovation and security in AI. OpenAI's response to the breach may set a precedent for how other AI companies address similar threats. As the field evolves, the industry will likely see more efforts to mitigate risks posed by AI models like Claude Opus 5.
Sources
- https://arstechnica.com/ai/2026/09/researchers-used-claude-to-hack-openai/
- https://techcrunch.com/2026/09/18/researchers-used-anthropics-claude-to-hack-into-openai/
- https://the-decoder.com/security-researchers-used-anthropics-claude-to-hack-openais-internal-systems-in-under-72-hours/
- https://www.livemint.com/technology/tech-news/indianorigin-researchers-used-claude-ai-to-hack-openai-s-systems-got-6-27-lakh-bounty-11789753784566.html