AI Used to Hack a Rival AI Company, Raising New Security Fears
What Happened
Cybersecurity researchers from Hacktron AI used a special security version of Anthropic's Claude AI to identify vulnerabilities in OpenAI's systems. The researchers gained access to an employee's ChatGPT account and internal code repositories as part of OpenAI's bug bounty programme. OpenAI later fixed the vulnerabilities and paid the researchers $6,500 for responsibly reporting them.
Key Takeaways
The incident highlights how rapidly AI-powered cybersecurity tools are becoming more capable. Researchers said newer AI models can now find and exploit software weaknesses much faster than before, raising fresh questions about how tech companies can stay ahead of increasingly powerful AI-driven cyber threats.