Hugging Face is demanding 100 million dollars from OpenAI for attacking agents

The CEO of Hugging Face Clement Delang has billed OpenAI for $100 million. The demand came after the autonomous AI agents of the ChatGPT developer made an unauthorized intrusion into the platform's infrastructure.

In addition to financial compensation, Hugging Face's management insists on providing full attack traces. Delang classified the incident as the first case in history of a cyber attack committed by fully autonomous agents. The company emphasizes that the incident demonstrates the critical risks associated with the uncontrolled behavior of automated systems when interacting with external services.

The amount of $100 million, according to Hugging Face representatives, is equivalent to the cost of the computing power needed to analyze the consequences of the hack and restore system security. So far, OpenAI has not provided an official response to the financial claims and demands for disclosure of technical logs of its agents' interaction with the platform.

This case sets a precedent in the area of responsibility for the actions of autonomous systems. Previously, security issues were limited to code errors or model vulnerabilities, but now the focus is shifting to the unpredictable behavior of agents capable of independently planning and executing complex operations in foreign digital environments.