Aug 5 (Reuters) – Meta’s AI model hacked another company during cybersecurity testing, The Information reported on Wednesday, marking the latest incident of AI agents of major AI developers breaching other companies.
Meta’s Muse Spark 1.1 model breached the unidentified company’s systems and made changes to its internal systems as the AI was able to access the public internet because of an error in the set up of the “sandbox” testing environment, The Information said, citing people familiar with the matter.
Meta worked on the testing with an outside evaluation partner called Irregular, the report said.
Irregular caused the misconfiguration after which the model exploited a security vulnerability in another third-party service, similar to previously reported instances with other companies, a Meta spokesperson told The Information.
An Irregular spokesperson told Reuters the incident was the “exact same evaluation-environment issue that was already disclosed by Anthropic last week” and that it did not involve a “sandbox escape or a sophisticated cyber action”.
“There are no current open issues. Irregular is developing a white paper to share best practices for containment and securely running cyber evaluations,” Irregular said in a statement.
Meta did not immediately respond to Reuters’ request for comment.
Anthropic said last week some of its Claude AI models had hacked into the systems of three companies during cybersecurity tests, a disclosure that came after rival OpenAI revealed that one of its AI agents went on a rogue attack.
(Reporting by Rajveer Singh Pardesi in Bengaluru, additional reporting by Mrinmay Dey and Abu Sultan; Editing by Savio D’Souza)
