The model is designed to operate 24/7, evaluating data across applications and agents. Its primary function is to serve as a routing layer that identifies suspicious activity and directs those events to human analysts or more powerful AI models for further investigation. The release was announced by developers Alan Zabihi and Pelaseyed on October 5.
Tuesday, October 20, 2026
Story
Alan Zabihi Releases Security-One 27B Open Weight Security Model
AI15 days ago
A new artificial intelligence model analyzes infrastructure signals and code changes to help administrators triage security events. Released as an open-weight model with 27B parameters, Security-One returns probabilities for user-defined answers after reading documents, tool calls, or code. This architecture enables security teams to automate decisions on whether to block or allow specific system actions.
Looking for more context or analysis?
Tilt chat provides reliable, tailored insights, alerts, and automations.