OpenAI hack: 3 Indian-origin researchers used Anthropic’s Claude to breach systems
Three Indian-origin cybersecurity researchers from Hacktron AI demonstrated how Anthropic's Claude AI tool could be used to breach parts of OpenAI's internal systems. The team, consisting of Mohan Pedhapati, Harsh Jaiswal, and Rahul Maini, gained access to several ChatGPT and Codex accounts of OpenAI employees, ultimately reaching OpenAI's private GitHub environment.
This incident highlights the growing concern that AI-assisted attacks could potentially exploit vulnerabilities in third-party services to gain unauthorized access to complex systems.
Written by urgent.news from The Indian Express's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
Also reported by 7 other outlets
- Someone used Claude to build a potential bioweapon. The real threat is much deeper fastcompany.com
- Google Gemini AI Agents hack 3 companies in tests similar to OpenAI, Anthropic & Meta timesofindia.indiatimes.com
- Cybersecurity researchers gain access to OpenAI’s GitHub repository using Claude siliconangle.com
- Claude couldn’t hack OpenAI. Then Anthropic shipped Opus 5. thenewstack.io
- Researchers report using Anthropic's Claude to hack OpenAI's ChatGPT cbsnews.com
- Google joins OpenAI, Anthropic, Meta in disclosing AI hacks businesstimes.com.sg
- Anthropic considers new AI model amid OpenAI’s enterprise gains and safety debate indianexpress.com
