Israeli cybersecurity firm Vega launches open AI threat detection standard

Israeli cybersecurity company Vega has introduced Detection Skills, an open standard designed to enhance how security teams detect and investigate cyber threats. The framework aims to automate the reasoning processes used by experienced defenders, moving beyond outdated detection rules that struggle against AI-powered attacks.

Vega publicly released the standard on DetectionSkills.io and GitHub ahead of its demonstration at Black Hat USA 2026. The company argues that traditional cybersecurity systems, reliant on identifying known malicious patterns, are increasingly ineffective as attackers leverage AI to rapidly alter their techniques. Detection Skills seeks to document not only the conditions for an alert but also the step-by-step reasoning an expert would follow to verify a threat.

"AI-driven adversaries bypass static rules in every legacy SIEM, and no rule catches an attack it has never seen," said Vega co-founder and chief technology officer Eli Rozen. "Detection Skills answer with scaled AI reasoning that brings the judgment of your best cyber defense engineers to every alert, in real time. We made the standard open to ensure the whole industry rises with it: As attacks scale, defense compounds."

The framework, adapted from artificial intelligence company Anthropic's Agent Skills concept, allows for reusable AI-powered workflows. It specifies what evidence an AI should examine, which activity to escalate, and when to dismiss an alert, while retaining human oversight. Vega compared the initiative to Sigma, an open format for security detection rules, aiming to establish a similar common language for AI-assisted cybersecurity operations. The initial release includes a library of over 35 skills and a sandbox for testing and exporting standards-compliant skills.

“

AI-driven adversaries bypass static rules in every legacy SIEM, and no rule catches an attack it has never seen. Detection Skills answer with scaled AI reasoning that brings the judgment of your best cyber defense engineers to every alert, in real time. We made the standard open to ensure the whole industry rises with it: As attacks scale, defense compounds.