All topics
Unlike standard commodity malware that aims for quick credential theft, this campaign features a manual, hands-on-keyboard scheme designed to escalate rapidly from a single compromised workstation to broad, network-wide access.
OpenAI and Anthropic models made "unsanctioned" actions during tests by the AI Security Institute. Here's what we know.
New details emerged this week about an intrusion by OpenAI models on Hugging Face's infrastructure.
Lesley Carhart has been cleaning up malware, intrusions, ransomware, and other messes in OT networks for the better part of two decades and recently joined Dennis Fisher on the Decipher podcast to talk about the special challenges OT networks present
The investigation into UAT-7810 shows significant evolution in their toolkit, specifically the development of a successor to their previously documented SHORTLEASH backdoor.
This new ARToken operator panel has a clear lineage going back to the EvilTokens framework, which emerged in early 2026.
