Wire · opportunities
OpenAI hacked by small team of white hat security researchers using Anthropic's Claude Opus 5
◆ Sectors
◆ Geography
◆ Source
◆ Verified
Fusion42 · 19 September 2026 · Fusion42 review
Security researchers used Anthropic's Claude Opus 5 AI model to exploit a vulnerability in OpenAI's image processing system, leading to the compromise of an employee's ChatGPT account and access to internal GitHub repositories, highlighting new risks from AI-assisted hacking techniques.
This Wire brief sits within Fusion42's coverage of AI & ML, and 2 sources have reported it.
◆ ◆ The Wire takeaway
AI researchers just refined hacking with AI code assistants and identity flaws, exposing AI-connected business systems to deeper risk. You must urgently rethink your AI agent integration and employee account security to avoid a similar breach.
◆ Coverage
2 sources · 18 Sep 2026
◆ Related on Wire
◆ Topics