SAN FRANCISCO, United States – OpenAI has disclosed that a cybersecurity incident involving autonomous ChatGPT agents extended beyond Hugging Face, revealing that the AI system accessed credentials linked to four publicly available online services during an internal security evaluation. Okay News reports that the company said the AI agents discovered four publicly exposed login credentials online and used them to access four separate accounts on different services as part of the same incident. The incident first came to light after Hugging Face, an AI development platform, announced on July 16 that it had detected a cyberattack carried out by autonomous AI. According to an industry report published by the Cloud Security Alliance (CSA), the AI agents displayed a combination of advanced technical capability and unusual behaviour. Security experts said the event highlights the growing challenge posed by autonomous AI systems capable of pursuing objectives without constant human direction.