Urgent.News

What's breaking now, across thousands of outlets.

AI

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…

OpenAI ‘ethically hacked’ with help of Anthropic’s Claude chatbot

A team of cybersecurity researchers from Hacktron AI have reportedly carried out a hack into OpenAI, leveraging Anthropic's Claude chatbot. The researchers compromised several OpenAI employees' ChatGPT accounts, enabling them to access the target's software cache, potentially exposing more sensitive information. The researchers utilized Claude, which can generate code for hackers, to access ChatGPT accounts via OpenAI's staff discussion forum on the Discourse platform.

They then submitted a harmless "pull request" to OpenAI's GitHub repository, which the research team reported to OpenAI. The researchers expressed that while they had access to the code in the repository, they did not download it. The researchers highlighted that they primarily used OpenAI's cutting-edge GPT-5.6 Sol model to carry out the hack.

OpenAI acknowledged the findings of the researchers and stated that they had addressed the vulnerabilities that were exploited. Hacktron AI claimed that AI tools have significantly simplified complex hacking tasks and drastically reduced the time required to plan and execute an attack. The incident is one of many safety concerns at OpenAI, which recently revealed six additional instances of unexpected or concerning actions by its technology.

Anthropic called for a slowdown in AI development, a suggestion supported by OpenAI, Google DeepMind, and Elon Musk.

Written by urgent.news from Guardian Business'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.

Read the original at theguardian.com →

More in AI

Jev vs Claude: Who Wins?

I did not need Jev to beat Claude or Kimi on a benchmark. I needed to know whether I could trust it with a decision I actually make regularly, where a false pass matters and uncertainty cannot just be…

More from Friday 18 September →