Report: Anthropic's Claude Helped Researchers Breach OpenAI Systems

An artificial intelligence model developed by Anthropic, called Claude, was reportedly used by cybersecurity researchers to breach systems belonging to rival company OpenAI, according to a report by Anadolu Agency.

The report cites the cybersecurity team Hacktron, which said it used Claude to gain access to OpenAI's internal code repository as part of a security assessment.

According to Hacktron, the team accessed the repository without viewing its contents before disclosing the vulnerabilities to OpenAI, the report said.

Details on the specific vulnerabilities identified, the scope of the assessment and OpenAI's response were not included in the available report.

The disclosure highlights the growing role of AI models in cybersecurity research, including efforts by one company's technology to probe the systems of a competitor in the same industry.

Anadolu Agency did not specify whether OpenAI or Anthropic issued separate statements confirming the details of the reported breach.