Asymmetric Security's agentic DFIR work featured in Google DeepMind's agent security paper
Google DeepMind's new white paper, The Three Layers of Agent Security: A Framework for Policymakers, highlights two efforts putting AI agents to work in security operations: Google's own Agentic Security Operations Center, and our work here at Asymmetric Security.
The paper is direct about the stakes. AI can "uplift attackers, increasing the speed, scale, and sophistication of cyberattacks that they can launch." Longer term, "more capable AI agents could fully automate complex malicious cyber operations."
By putting AI agents to work in security operations, defense can scale as fast as the attacks do.
Google's Agentic Security Operations Center builds AI agents that run defensive security operations workflows on behalf of defenders.
At Asymmetric Security, we build agentic systems for digital forensics and incident response, transforming cyber forensic investigations into a scalable process that can take hours instead of days.
Cybersecurity is entering its biggest transformation in decades, and our mission is to help make it go well. We're glad to see our work recognized in the paper, alongside colleagues from Mandiant (part of Google Cloud), Wiz, XBOW, and elsewhere.
Read the paper: The Three Layers of Agent Security