Asymmetric Security Announces its Launch with a Mission to Accelerate AI Cyberdefense

Asymmetric Security, the first AI-native Digital Forensics and incident response company, led by Zainab Ali Majid, Pippa Thompson, and Alexis Carlier, has announced its launch and raised a $4.2M pre-seed led by Chad Byers at Susa Ventures, with participation from others including Halcyon Ventures, Overlook Ventures, Seldon Lab, Matt Clifford, Geoff Ralston, Charlie Songhurst, and AI and security leaders from Anthropic and Google DeepMind.

Hundreds of real-world cyberattacks have been mitigated using internally developed AI systems. Insights from these incidents are used to build evaluations that strengthen AI cyberdefense capabilities across frontier AI labs.

Modern security stacks excel at broad scanning but struggle with depth. They generate vast volumes of alerts, most of which are never fully investigated. When serious threats slip through, organizations rely on a scarce and expensive resource: human digital forensics experts, typically deployed only after a confirmed breach.

Cyber risk lives in the gap between surface-level detection and deep understanding.

This gap narrows in a world of abundant intelligence. AI agents do not merely automate existing workflows; they unlock entirely new categories of work that were previously too slow or costly to perform at scale.

In cyberdefense, one of the most significant shifts will be the emergence of automated digital forensics as a core detection mechanism. Once automated, digital forensics no longer needs to be reactive or rare; it can become the default standard for identifying threats.

Software is, for the first time, displacing services. Full-stack AI-driven DFIR is emerging as a venture-scale business, with forensic investigations, once the exclusive domain of human experts, now automatable and scalable like software.

This shift also creates durable technical and distribution advantages as digital forensics evolves into a core method of threat detection.

During the transition to cloud computing, CrowdStrike began as a DFIR services firm and used that wedge to reinvent threat detection for the cloud era, ultimately becoming a category-defining cybersecurity company.

DFIR services represent an even stronger wedge in the AI era. Real-world incident response enables the creation of high-fidelity evaluations and expert reasoning traces for reinforcement learning, while tight feedback loops emerge as internal analysts actively use and stress-test AI systems.

Incident response also builds deep trust with enterprises at their most critical moments. In cybersecurity, trust drives distribution, and the moment immediately following a crisis is when adoption decisions are made.