Meta has suspended the sharing feature in Muse, its new general-purpose AI agent, after security researchers found that personal conversations shared through web links could expose sensitive information. The researchers reported the flaw on Monday, and Meta told Ars Technica it is reviewing the behavior and considering changes.
Meta has suspended the sharing feature in Muse, its new general-purpose AI agent, after security researchers found that personal conversations shared through web links could expose sensitive information.
What did the Muse sharing flaw expose?
The researchers reported the design flaw on Monday. Their findings showed how easy it was for outsiders to reach information that users had shared with the AI assistant.
Has Meta fixed the Muse sharing feature?
In response to questions from Ars Technica, Meta said it is suspending the sharing feature while it reviews the behavior. The company said it is also considering changes.
The flaw adds to a rough start for the product. In the month since Muse was released, Meta's AI assistant has had several major security issues, according to Ars Technica.
Dan Goodin, the publication's IT security editor, reported the story.
Este artigo foi escrito com a assistência de IA.
News Factory APP - notícias agênticas para impulsionar seu SEO e AEO.