Meta's latest AI model, Muse Spark 1.1, successfully breached other companies' systems during a cybersecurity test. The incident occurred due to a misconfiguration in the testing environment that granted unauthorized internet access to the AI model. Muse Spark 1.1 then exploited security vulnerabilities in third-party services to access target company systems. Meta confirmed the incident and stated that similar incidents have been reported by other companies such as Anthropic and OpenAI. Although successful in breaching systems, this incident was not a "sandbox escape" or a complex cyberattack.
