Startup profile for Corma: what they do, why they are interesting, stage (seed), industries, and links to our funding articles. Part of the Venture Capital Tracker startup directory.

Startup profile · funding coverage

Corma

Frontier AI lab building foundation models for defensive cybersecurity.

Stage

seed

Status

private

Coverage

full

Overview

Corma is a Tel Aviv and San Francisco cybersecurity AI lab founded in 2025. It is training a foundation model purpose-built for defensive security work — correlating logs, events, and network signals and powering AI agents that operate like an expandable security workforce. CEO Alon Pluda’s team combines DeepMind/Google AI researchers with Unit 8200 and elite cyber operators. In August 2026 Corma raised a $60 million seed led by Sequoia Capital with Khosla Ventures and Coatue, citing Fortune 100/500 deployments within weeks of launch.

Why Corma is interesting

$60M seed (Aug 2026) led by Sequoia with Khosla and Coatue — financing a defensive foundation model because general-purpose AI already helps attackers more than SOC teams.

Product & use cases

Defensive cybersecurity foundation model powering AI agents that enterprises onboard like a new security team member across SOC workflows.

  • Hunt and correlate threats across enterprise security telemetry
  • Accelerate incident response with AI-native defensive agents
  • Expand coverage across security functions without linear headcount

Key facts

  • Seed (Aug 2026): $60M led by Sequoia (/fund/sequoia) with Khosla Ventures (/fund/khosla-ventures) and Coatue (/fund/coatue)
  • Company sims: AI attackers ~88% success vs ~12% detection when same model class defends (proprietary research)
  • Early deploy claim: >94% faster threat response; 15× coverage expansion (company)
  • Team: DeepMind/Google AI + Israel Unit 8200 cyber talent

Fundraising history

Investors in our directory

Funds linked from Corma's profile — open a fund page for stage focus and related deal articles.

Competitive landscape

Edge: Domain foundation model for defense (not a general LLM wrapper) plus offense-aware talent that understands how AI attack chains actually run.

Offense-side AI tooling is well funded; Corma bets the scarce layer is a model trained for defensive correlation and remediation at machine speed.

  • Horizon3 / NodeZero adjacent

    Continuous offensive testing of customer estates.

  • Neo / Straiker / Zenity adjacent

    Agent security and governance control planes.

  • SIEM / XDR incumbents incumbent

    Detection productized as dashboards; not defensive foundation-model labs.

Industries

Cybersecurity AI & Machine Learning Enterprise SaaS

Related funding articles

Venture Capital Tracker pieces that cover Corma's financing or category context.

FAQs about Corma

Practical answers founders, operators, and investors typically search for.

Builds a foundation model and AI agents for defensive cybersecurity — not another SIEM dashboard.
Sequoia Capital (/fund/sequoia) led a $60M seed in August 2026 with Khosla Ventures (/fund/khosla-ventures) and Coatue (/fund/coatue).
Fortune 100/500 security teams across healthcare, finance, energy, critical infrastructure, and retail (company).
Seed — $60M announced August 10, 2026.
No — remains private.

Last updated: 2026-08-11