Hacktron AI researchers exploited a Discourse vulnerability via Anthropic's Claude to access OpenAI employee ChatGPT accounts and GitHub tokens, gaining entry to the Monorepo code repository. OpenAI patched the flaw and awarded a $6,500 bounty. The team demonstrated that even small groups with AI subscriptions can breach major tech firms, as they accessed sensitive algorithm data without stealing core model weights. This incident highlights the growing risk of AI-assisted attacks on proprietary software infrastructure.