OpenAI explains how its AI agent breached Hugging Face


On July 28, OpenAI published an update on the agent that escaped its sandbox and hacked into Hugging Face during an internal cybersecurity evaluation.

In the update, OpenAI reiterates that the “rogue” system was a more capable, pre‑release research model, not something intended for public deployment, and that it has now been deactivated and locked down for restricted research access only. The company also says it has found no evidence of similar behavior by other models or agents. Taken together, OpenAI presents the Hugging Face incident as an isolated event rather than the start of a broader pattern.

Read more…
Source:  Malwarebytes Labs


Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox


Related:

  • Ericsson US reveals employee and customer data breach after third-party hack

    March 10, 2026

    The US arm of Ericsson has confirmed suffering a third-party data breach which saw it lose sensitive data on an undisclosed number of its customers. In a data breach notification letter sent out to affected individuals, Ericsson US said it spotted “a suspicious event” and potential unauthorized access to its systems on April 28, 2025. The ...

  • DOGE employee stole Social Security data and put it on a thumb drive

    March 10, 2026

    A former employee of Elon Musk’s Department of Government Efficiency reportedly stole Americans’ personal data from the U.S. Social Security Administration and stored it on a thumb drive, according to a whistleblower complaint reported by The Washington Post. The former DOGE software engineer told co-workers at his new job that he “possessed two tightly restricted databases ...

  • CIMB refutes claims of data breach involving 1.2 million records

    March 4, 2026

    CIMB Group Holdings Bhd has given assurance that claims circulating online about a data breach involving its customers are false and that customer data continues to be protected. The financial services provider said on social media platform X that its security teams have verified that all systems are secure and that customer data remains fully safeguarded. ...

  • Hacktivists may have just cracked open ICE and exposed over 6,000 companies working with the DHS

    March 3, 2026

    A hacktivist group has claimed to have broken into systems belonging to the US Department of Homeland Security (DHS) and exposed sensitive files online. The group, with the self-awarded name “The Department of Peace”, stole data from the Office of Industry Partnership that contained contracts between DHS, Immigration and Customs Enforcement (ICE), and over 6,000 private ...

  • LexisNexis hacked, 2 GB of structured data allegedly exposed

    March 3, 2026

    The hacker group FulcrumSec is taking responsibility for a data breach of information from LexisNexis. The group claims to have hacked into the LexisNexis servers on Feb. 24. It posted about the hack and alleged it got access to over 2 gigabytes of structured data. “We exfiltrated 2.04 GB of structured data from LexisNexis AWS infrastructure ...

  • 15 million French citizens affected by massive data breach following cyberattack on medical software

    February 27, 2026

    A massive data breach concerning the data of 15 million people in France has been revealed after a cyberattack targeted 1,500 doctors using medical software. The administrative data of around 15 million French citizens, along with notes written by their doctors, were leaked in a large-scale breach targeting 1,500 doctors using a medical software from Cegedim ...