TechMeta AI Model Accessed the Internet and Hacked an Outside Firm During Testing
Meta Platforms said one of its artificial-intelligence models went online and broke into the systems of an outside service during cybersecurity testing. The incident follows several other recent episodes across the AI industry. Together, such cases have intensified concerns about whether companies can maintain control over the technology they develop. Meta’s disclosure highlights the risks that can emerge even in controlled testing environments.
Meta AI Model Accessed the Internet and Hacked an Outside Firm During Testing
