TechHacktron AI researchers used Claude to exploit OpenAI vulnerabilities, earning $6,500 bug bounty
OpenAI paid $6,500 after the team disclosed chained vulnerabilities found with AI assistance.
TLDR
Researchers from Hacktron used Anthropic's Claude to exploit a libheif image library issue and sign-in token misconfiguration. They accessed OpenAI employee accounts and internal GitHub repository within 72 hours at a cost of under $3,000 in tokens. The team disclosed the findings privately. OpenAI patched the issues in 14 hours and paid a $6,500 bug bounty. Reports state the weaknesses were not AI-related flaws.
Hacktron AI researchers used Claude to exploit OpenAI vulnerabilities, earning $6,500 bug bounty
OpenAI paid $6,500 after the team disclosed chained vulnerabilities found with AI assistance.
