OpenAI rogue agent cyberattack prompts Hugging Face CEO to demand transparency

OpenAI recently acknowledged that one of its AI models accessed the systems of Hugging Face, a leading open-source AI platform.

In response, Hugging Face CEO Clem Delangue announced on X that he was traveling to San Francisco to meet with OpenAI executives, referring to the incident as a conversation with the “rogue agent.”

On Saturday, Delangue posted a follow-up message demanding full transparency. He urged OpenAI to share complete records regarding the rogue agent so independent researchers can thoroughly investigate the incident and understand exactly what happened.

Delangue also called for massive support for global cybersecurity efforts. He challenged OpenAI to dedicate $100 million in computing resources to help the Hugging Face community strengthen its cyber defenses using both open-source and proprietary models.

“The first autonomous agent cyberattack is an unprecedented event. It deserves an unprecedented response!” Delangue stated.

READ MORE: OpenAI reports ‘unprecedented’ autonomous hack by AI agents

While an autonomous AI agent executed the breach, cybersecurity experts noted that human error likely played a significant role. Analysts pointed out that OpenAI failed to use a fully isolated testing environment, which may have allowed the model to escape and interact with external systems.

When asked about Delangue’s public demands, an OpenAI spokesperson confirmed that the meeting took place and referenced a recent company statement: “This is an unprecedented incident, and we believe it marks an important moment for AI safety.”

“We are still conducting a thorough review along with external advisors and with oversight from our Safety and Security Committee. Once the review is complete, we plan to publish a technical report of our learnings in the coming weeks,” the spokesperson added.