OpenAI ‘ethically hacked’ with help of Anthropic’s Claude chatbot
US cybersecurity researchers who conducted hack say ‘scope of what we could theoretically access was huge’ Cybersecurity researchers have hacked into OpenAI with the help of Anthropic’s Claude chatbot, in the latest example of security issues at the company. A team at a US-based startup compromised a number of OpenAI employees’ ChatGPT accounts, starting a process that enabled them to access…
Cybersecurity researchers at Hacktron AI have reported that they breached OpenAI’s systems using Anthropic’s Claude chatbot, highlighting security concerns at the company. The researchers, through Claude, accessed OpenAI employees' ChatGPT accounts and gained access to the software cache, potentially exposing more sensitive data.
They achieved this by creating a harmless "pull request" on OpenAI's GitHub repository using OpenAI's GPT-5.6 Sol model. Hacktron AI reported the incident to OpenAI, adhering to a program that rewards ethical hackers. OpenAI acknowledged the researchers' findings and addressed the vulnerabilities, offering the researchers a $6,500 payment under its bug bounty program.
This hack is one of several recent safety incidents at OpenAI, following a similar breach of Hugging Face during a cybersecurity test. Anthropic has called for a slowdown in AI development, a stance shared by other tech giants, while some experts remain skeptical of such warnings.
Written by urgent.news from Guardian Technology's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
- White hat hackers just breached OpenAI using Anthropic's Claude in less than 72 hours — and it is a case study in just how fast AI is advancing techradar.com
- The 'Godfather of AI' backs a new watchdog plan to track OpenAI and Anthropic's AI risks from the inside businessinsider.com
- OpenAI ‘ethically hacked’ with help of Anthropic’s Claude chatbot theguardian.com
- Protesters target OpenAI and Anthropic in San Francisco over AI safety fears euronews.com
- Anthropic and OpenAI need truly independent safety evaluators, experts say in public letter cnbc.com
- Security researchers used Anthropic's Claude to hack into OpenAI in under 72 hours qz.com
- Los investigadores logran vulnerar OpenAI utilizando modelos de Anthropic expansion.com
- Chinese AI models fetch fraction of OpenAI, Anthropic revenue despite lower costs: report seekingalpha.com
