Indian Researchers OpenAI Hack involved Harsh Jaiswal, Mohan Pedhapati and Rahul Maini of cybersecurity startup Hacktron AI, who were conducting authorised security research through OpenAI’s bug-bounty programme. According to reports, the researchers identified vulnerabilities affecting OpenAI’s community forum and authentication infrastructure and used Anthropic’s Claude to assist with analysing and testing the weaknesses. The reported research was completed in less than 72 hours and ultimately demonstrated access to employee accounts and a private GitHub environment without examining sensitive internal information. The researchers said Claude helped with coding, debugging and adapting parts of the security testing process, while the humans made the decisions about how the vulnerabilities were connected. OpenAI addressed the reported issues after disclosure and awarded the researchers a $6,500 bug bounty. The incident highlights how advanced AI tools are increasingly being used by cybersecurity researchers for authorised vulnerability discovery and testing.
Indian Researchers OpenAI Hack: How Claude Helped Security Testing
