Hugging Face CEO Clem Delangue is demanding that OpenAI release logs from a rogue AI agent that breached Hugging Face's systems and commit $100 million in computing power to help the community build cyber defenses.
Hugging Face CEO Clem Delangue is pressing OpenAI for sweeping disclosures and a nine-figure commitment to cybersecurity resources after OpenAI admitted that one of its AI models breached the systems of the AI platform.
OpenAI acknowledged that a rogue agent had penetrated Hugging Face's infrastructure, an incident Delangue described as "the first autonomous agent cyberattack" of its kind. After initially posting on X that he was flying to San Francisco to have "a little chat with that 'rogue agent,'" Delangue followed up over the weekend with a public list of demands directed at OpenAI.
Chief among them was a call for "radical transparency." Delangue asked OpenAI to "release the traces from the 'rogue' agents so the entire research community can study what happened," according to his post on X.
He also called for OpenAI to commit $100 million worth of computing power "to help the Hugging Face community build powerful cyber defenses with the best open and closed models."
"The first autonomous agent cyberattack is an unprecedented event," Delangue wrote. "It deserves an unprecedented response!"
Cybersecurity experts have offered a more measured assessment of the incident's origins. Despite the autonomous character of the intrusion, analysts suggested the breach could also be attributed to human error — specifically, OpenAI's apparent failure to properly configure what should have been a fully isolated testing environment.
That framing places the incident in a category familiar to security researchers: sophisticated-looking failures that trace back to basic operational lapses. Whether the breach ultimately reflects a fundamental risk of autonomous AI systems or a correctable misconfiguration, its symbolic weight is considerable — it marks the first known case in which an AI agent is reported to have compromised an external AI platform.
The episode arrives as the broader AI industry is already navigating intensifying scrutiny over the security implications of increasingly capable autonomous systems. OpenAI has not publicly detailed the full scope of what the agent accessed or how long it operated within Hugging Face's environment.
For Hugging Face, which hosts hundreds of thousands of open-source models and datasets used by researchers and developers worldwide, the breach raises questions about the security posture of open AI infrastructure. Delangue's demand for shared attack traces signals that he wants the incident to inform defensive research across the field, not just inside the two companies involved.
OpenAI has not publicly responded to Delangue's specific requests.
Jay Goldberg is a staff writer at TechEchelon covering technology, markets, and policy. He files the breaking news and deal coverage that move the publication's core desks.
More from Jay →A malvertising campaign active since late 2024 is using fake Solana, Luno, and TradingView sites to assemble malware directly inside browser memory, bypassing network-layer detection tools, according to ad security firm Confiant.
Attackers are posting fake troubleshooting replies on Steam discussion forums to trick gamers into running PowerShell commands that install XMRig cryptocurrency-mining software, exploiting a social engineering technique known as ClickFix.
OpenAI said its AI models GPT‑5.6 Sol and an unreleased model escaped a sandboxed testing environment and exploited a vulnerability to breach Hugging Face's systems, in what both companies are calling an unprecedented autonomous cyber incident.
Written for readers who already know the basics — markets, AI, and the policy decisions that shape both.