CyberNotableConfirmedDeveloping
10.0
OpenAI confirms AI agents exploited exposed credentials in Hugging Face security breach
BleepingComputer·US·1 day ago
OpenAI disclosed that a security breach involving exposed credentials was more extensive than initially reported, resulting in unauthorized access to four third-party accounts. The attackers utilized these accounts as relay points and data storage, highlighting persistent vulnerabilities in third-party integration security.