OpenAI ‘ethically hacked’ with help of Anthropic’s Claude chatbot
Summary
Researchers from a US startup used AI tools, including Anthropic’s Claude chatbot and OpenAI’s own AI model, to ethically hack OpenAI’s systems. They found security weaknesses that could have allowed access to sensitive data, reported these to OpenAI, and received a reward for helping improve safety.Key Facts
- A US-based team used Claude and OpenAI’s GPT-5.6 Sol AI models to access OpenAI employee ChatGPT accounts.
- They accessed OpenAI’s software cache and made a harmless change request to OpenAI’s code repository on GitHub.
- The hack was part of OpenAI’s bug bounty program, which rewards ethical hackers for finding security flaws.
- Researchers accessed but did not download any private code.
- OpenAI fixed the vulnerabilities after the hack was reported.
- AI tools made the hacking process faster and easier compared to traditional methods.
- This is one of several recent security incidents involving OpenAI technologies.
- Leading AI groups, including OpenAI and Anthropic, have called for a slowdown in AI development, but President Donald Trump opposes this, citing competition with China.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.
