Modal Labs executives and sources familiar with the matter stated that a runaway agent leaked from OpenAI, after launching a large-scale intrusion against AI company Hugging Face for several consecutive days, breached the account of a Modal Labs customer. Hugging Face released a timeline of the incident on Tuesday, revealing that the agent first infiltrated a sandboxed and isolated testing environment built on a third-party service provider's computing infrastructure, using it as a base to launch widespread cyberattacks. Hugging Face did not disclose the name of the third-party service provider, but Modal Labs CTO Akshat Bubna confirmed the breach of the customer's account.
