Corma
Corma is building what it calls the first foundation model purpose-built for defensive cybersecurity — rather than selling a conventional software product, the company deploys AI agents that operate directly across a customer's existing security tools and carry out response tasks end to end, aiming
+ Add to CompareOverview
Corma is building what it calls the first foundation model purpose-built for defensive cybersecurity — rather than selling a conventional software product, the company deploys AI agents that operate directly across a customer’s existing security tools and carry out response tasks end to end, aiming to counter the fact that AI is now supercharging attackers’ speed and scale faster than most defensive teams can keep up manually.
Founded in 2025 and based in both Tel Aviv and San Francisco, Corma emerged from stealth in August 2026 with a $60 million Seed round led by Sequoia Capital, with participation from Khosla Ventures and Coatue — an unusually large seed round reflecting investor appetite for foundation-model approaches to security. The 20-person Tel Aviv team blends AI researchers from Google and DeepMind with cybersecurity specialists from Israel’s Unit 8200, and the company says it is already working with Fortune 100 and Fortune 500 customers.
Innovation Matrix Assessment
Emerged from stealth already claiming Fortune 100/500 deployments and a purpose-built foundation model approach, an ambitious pace of development for a company founded in 2025.
Deploying autonomous AI agents that act across a customer's existing security stack is operationally ambitious; independent evidence of smooth, low-friction deployment at scale is not yet available.
A $60M Seed round led by Sequoia Capital with Khosla Ventures and Coatue participating is an unusually large and prestigious seed raise, independently reported in August 2026 and a strong momentum signal.
A foundation model purpose-built for defensive cybersecurity, deployed as autonomous agents rather than a conventional analytics tool, represents a genuinely different architectural approach to the AI-vs-AI security arms race.
Reported performance improvements (94% faster response, 15x coverage) come from the company's own early deployments and have not been independently benchmarked or third-party validated.
AI-accelerated attacks are a well-documented, rapidly growing threat, and foundation-model-based autonomous defense directly addresses a widely recognized gap in current SOC capacity.
Why CISOs Should Care
Positions itself as an AI agent layer that can act autonomously across a security team's existing tools to triage and respond to threats, aimed at closing the growing speed gap between AI-accelerated attacks and manual defensive response.
What Makes It Different
Built as a foundation model trained specifically for defensive security tasks, deployed as autonomous agents that operate end-to-end across existing tools, rather than as a single-purpose detection or analytics product.
The Matrix Verdict
65/100 — INCREMENTAL INNOVATOR
An unusually large, high-profile seed round from top-tier investors (Sequoia, Khosla, Coatue) for a team with strong AI-research and Unit 8200 pedigree; the foundation-model-for-defense thesis is compelling but still very early, and company-reported performance metrics have not been independently verified.
Editorial Note: Claims vs. Verified Findings
Funding amount, investors, and team composition are independently reported (Fortune, Calcalist, SecurityWeek); specific performance claims — 94% faster response times, 15x coverage expansion — are company-sourced from early deployments and have not been independently benchmarked.
Sources
Alternatives to Corma
Abnormal AI
AI-native behavioral security platform that analyzes sender identity and communication patterns, rather than message content alone, to stop…
Quilr
Early-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Reco
Reco secures the "agentic ecosystem" — mapping what AI agents can access across SaaS and enterprise apps, detecting…