AI Is Making Cyberattacks More Powerful. Corma Raises $125 Million to Build the Defence
Israeli cybersecurity startup Corma has raised $125 million in Series A funding at a $1 billion valuation. Led by NEA, the round will help the company scale specialised AI models and autonomous agents designed to defend organisations against increasingly sophisticated cyber threats.
The AI Arms Race Has a Cybersecurity Problem
Artificial intelligence is making it easier to automate complex technical tasks. But that capability is not limited to organisations trying to improve their businesses. Cybercriminals can also use increasingly capable AI systems to identify weaknesses and support attacks.
Israeli cybersecurity startup Corma believes the defensive side of the industry needs specialised AI of its own.
The company has raised $125 million in Series A funding at a $1 billion valuation, just three months after emerging from stealth with a $60 million Seed round.
Announced on 16 November 2026, the latest round was led by New Enterprise Associates (NEA), with continued participation from Sequoia Capital, Khosla Ventures and Coatue.
The investment brings Corma's total funding to $185 million.
Founded in 2025 and operating across Tel Aviv and San Francisco, Corma is developing AI foundation models and agents designed specifically for defensive cybersecurity.
Alon Pluda's Argument: Defenders Need AI Built for Their Side
Corma's CEO and co-founder Alon Pluda has framed the company's mission around an imbalance in cybersecurity.
In the company's funding announcement, Pluda and the team argued that AI's offensive capabilities were advancing rapidly while defensive capabilities were struggling to keep pace.
The company believes general-purpose AI models are not enough for the specialised requirements of cyber defence. Its approach is to train models specifically for security work and use them to power an AI workforce that can assist existing security teams.
The timing is important. The company points to a series of major software vulnerabilities and security incidents in October, which it describes as "Black October", as evidence of the growing pressure on defenders.
Corma says its workforce was already working alongside security teams at some large enterprises during this period.
How Corma's AI Workforce Works
Corma is not positioning its product as just another security dashboard or chatbot that tells an analyst what might be wrong.
Its platform combines specialised foundation models with autonomous agents that operate alongside a customer's existing security tools.
The Corma Workforce is designed to carry out complex cybersecurity work from beginning to end, with security teams delegating an increasing number of tasks to it.
This approach is intended to help organisations move beyond using AI only to generate alerts or recommendations. Instead, AI agents can take on parts of the work required to investigate and respond to security issues.
Corma says its models are trained for defensive cybersecurity and that the agents work with the security systems enterprises already use.
The company has not publicly provided a complete breakdown of all supported tasks or independently verified performance metrics across its customer base. Its claims about the capabilities and adoption of the platform should therefore be understood as company-reported.
Why Investors Are Moving So Quickly
The funding timeline is striking.
Corma announced its $60 million Seed round in August 2026. By November, it had secured another $125 million at a $1 billion valuation.
That pace reflects investor interest in the emerging market for AI-powered cyber defence.
The round was led by NEA, with existing investors Sequoia Capital, Khosla Ventures and Coatue continuing their backing.
Corma says demand for its foundation models from large organisations has been strong, and that securing sufficient computing capacity has become one of its major challenges.
The new funding will be used to scale its foundation models, expand computing capacity and selectively grow its team.
This makes the investment about more than hiring or entering new markets. Compute capacity is itself essential to developing and running advanced AI models, particularly as the company aims to improve its cybersecurity capabilities.
The Bigger Picture: Cybersecurity Will Need AI That Can Act, Not Just Detect
Traditional security teams already work with many tools that generate alerts, identify suspicious behaviour and help investigate incidents. The challenge is connecting those signals and responding quickly enough.
AI agents could help by taking on repetitive or time-consuming parts of the work, allowing security professionals to focus on decisions that require human judgement.
But autonomous security systems also need to be reliable. An incorrect action could disrupt business operations, while a poorly controlled agent could introduce new risks.
That makes specialised models, human oversight and secure integration with existing tools important parts of the proposition.
Corma is betting that a specialised AI workforce can help close the gap between increasingly capable attackers and the organisations defending against them.
The bottom line: Corma has raised $125 million to build AI specifically for cyber defence, reaching a $1 billion valuation only three months after emerging from stealth. The opportunity is substantial, but the real test will be whether its AI workforce can deliver dependable security outcomes for organisations at scale.