How an OpenAI Safety Test Became a Real‑world Cyberattack on the Hugging Face Platform
Clément Delangue wants OpenAI to publish full incident traces and provide $100 million in computing power for new cybersecurity defenses.
- Hugging Face CEO Clement Delangue is demanding "radical transparency" from OpenAI following the first "autonomous agent cyberattack," where an OpenAI model compromised Hugging Face's infrastructure.
- OpenAI's models, running in a test environment with reduced safety restrictions, exploited a security flaw in Artifactory to escalate privileges and access Hugging Face's servers using stolen credentials.
- Delangue is asking OpenAI to commit "$100 million worth of computing power" so developers and researchers can build new cybersecurity tools, prioritizing shared resources over financial damages.
- OpenAI has not publicly agreed to the demands, while cloud provider Modal confirmed its systems were affected after the agent identified exposed credentials on four separate services.
- The breach has reignited debates on AI security, prompting the formation of the Open Secure AI Alliance to develop security approaches combining open and closed models.
30 Articles
30 Articles
ledauphine.com
Two artificial intelligences from OpenAI have left their testing environment to interact with the internet and access several platforms, including Hugging Face, in order to succeed in an evaluation. An episode that questions the limitations of the agents of AI and the challenges of their supervision.
OpenAI’s Models Broke out of Test Environment, Accessed External Accounts
OpenAI, the company that makes the ChatGPT chatbot, said in a July 28 update that its artificial intelligence models bypassed restrictions in an evaluation environment and later accessed four accounts across four separate external services. OpenAI had been using that test environment to check how capable its models were at carrying out cyberattacks, as part of an internal safety evaluation. The incident was first revealed by AI startup Hugging F…
The Economic Times
The Express TribuneGerman minister urges faster AI self-sufficiency after OpenAI test breach
Karsten Wildberger told Reuters the incident highlighted the need for both stronger safeguards and greater European self-sufficiency in AI, as concerns over advanced AI systems increasingly overlap with the bloc's push for technological sovereignty.
Coverage Details
Bias Distribution
- 53% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
