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:
- Smiths Group: Shares fall as engineering giant hit by cyber attack
January 28, 2025
Global engineering firm Smiths Group has reported a cyber security incident involving unauthorised access to its systems. Upon detecting the breach, the firm promptly isolated the affected systems and activated its business continuity plans to mitigate disruptions. The company, known for its baggage screening equipment and explosive detectors, is collaborating with cyber-security experts to restore the ...
- The Honeymoon for Cloud Services Is Over
January 27, 2025
The cloud services you rely on are no longer as secure as they used to be. Once seemingly a safe haven for data and applications, attackers are increasingly leveraging cloud services for command and control—and the Symantec Threat Hunter Team predicts an unnerving upshoot in 2025. The Microsoft breach by Russian nation-state actors is one instance ...
- TalkTalk investigating data breach after hacker claims theft of customer data
January 27, 2025
U.K. telecoms giant TalkTalk has confirmed that it is investigating a data breach after a hacker claimed to have stolen the personal information of millions of customers. In a post on a popular cybercrime forum seen by TechCrunch, an individual using the alias “b0nd” claimed to have stolen the personal data of more than 18.8 ...
- FBI Agents’ Call And Text Logs Potentially Stolen In Data Breach
January 23, 2025
The FBI has raised alarm that hackers who breached AT&T’s system last year may have stolen months of agents’ call and text logs, which could potentially lead to the identities of anonymous informants connected to investigations. While the hackers did not access the content of conversations, the stolen call log metadata—records of who called whom, when ...
- eCommerce data breach exposes details on half a million users
January 21, 2025
The North Pole Company, a Canadian gift basket delivery service, allegedly suffered a data breach in which half a million customers lost sensitive personal information. The claim was made on BreachForums, a popular underground community where cybercriminals come to share tools, resources, and experiences, to find partners and plan future attacks. As cybersecurity researchers from Incogni ...
- HPE’s sensitive data exposed in alleged IntelBroker hack
January 20, 2025
IntelBroker has struck again. This time, the notorious BreachForums bigwig, which has a long list of high-profile victims, including Europol, Cisco, and GE, has claimed to have breached IT giant Hewlett Packard Enterprise (HPE). The suspected Serbian-origin hacker is offering to sell on BreachForums, sensitive data allegedly stolen from HPE including product source codes and personally ...