Skip to content
News on crypto, fintech and digital infrastructure

News Funding

Corma Funding Reaches $125 Million Series A Led by NEA for Defensive Cybersecurity Models

Cybersecurity startup Corma announced a $125 million Series A on November 16, 2026, led by NEA and valuing the company at $1 billion.

Rating 4.7 14 votes
Editorial 3D graphic showing $125M Series A funding for Corma defensive cybersecurity infrastructure.

Corma announced a $125 million Series A funding round on November 16, 2026. The round was led by NEA with participation from Sequoia Capital, Khosla Ventures and Coatue. The investment values Corma at $1 billion and brings total funding to $185 million.

Corma Funding Round Details and Valuation

The round follows a $60 million seed round completed in August 2026. Corma was founded in 2025 with headquarters in Tel Aviv and San Francisco. It emerged from stealth in August 2026 with a proprietary foundation model built specifically for defensive cybersecurity tasks.

CEO Alon Pluda said the surge in AI-driven attacks during October 2026 made clear that cybersecurity has become a contest between intelligent systems. The company reported strong demand from large enterprises, which has required rapid expansion of compute resources. Since August the firm increased server capacity 24-fold yet still maintains a waitlist for new customers.

Proceeds will support further scaling of foundation model training, additional compute infrastructure and selective hiring. The team currently numbers between 20 and 25 people with expertise drawn from foundation model research and cybersecurity operations, including veterans of Unit 8200.

Foundation Models and Autonomous AI Agents for Defense

Corma trains its own foundation models rather than fine-tuning general-purpose large language models. These models power a system called Corma Workforce that deploys autonomous AI agents inside customer environments. The agents connect to existing security tools, ingest logs, events and network telemetry, detect threats, correlate signals across data sources and support response actions.

Isometric infographic depicting autonomous defensive AI agents monitoring an enterprise cybersecurity grid.
Autonomous AI agents actively neutralize threats across enterprise network architecture.

Internal benchmarks show the models outperform general systems such as GPT-6 Astra and Claude Opus 5.5 on the DefenseBench V5 evaluation focused on defensive tasks. The company states certain breach response workflows can be accelerated by up to 100 times compared with manual processes. Current customers include Fortune 100, Fortune 500 and Global 500 organizations in finance, healthcare, energy and critical infrastructure. Annual recurring revenue is estimated in the tens of millions of dollars.

In tests cited by the company, general models succeeded in roughly 85 percent of simulated attacks but detected only about 12 percent of planted threats. Corma positions its specialized models as a direct response to this performance gap exposed during the wave of incidents in October 2026.

Market Context and Next Development Steps

The funding arrives amid documented growth in AI-assisted attacks and a cluster of high-impact vulnerabilities and breaches during October 2026. Industry observers have noted an imbalance between offensive and defensive AI capabilities, which Corma aims to address through purpose-built models.

The company plans to continue model development toward higher levels of capability for defensive cybersecurity. No specific release dates or additional milestones have been disclosed. Organizations interested in the platform remain subject to availability while compute and engineering resources are scaled.

Frequently asked questions

When did Corma raise its Series A round?

Corma announced the $125 million Series A on November 16, 2026. The round was led by NEA and followed a $60 million seed round completed three months earlier.

What does Corma use its funding for?

The proceeds will accelerate training of specialized foundation models, expand compute capacity and support limited team growth. The company has already increased server resources 24-fold since August 2026.

How do Corma models differ from general large language models?

Corma trains its own foundation models from the ground up for defensive cybersecurity rather than fine-tuning existing general models. The resulting agents integrate with customer security tools to analyze telemetry and support threat response.

Which sectors use Corma technology?

Corma serves Fortune 100 and Global 500 organizations across finance, healthcare, energy and critical infrastructure. Annual recurring revenue is estimated in the tens of millions of dollars.

Comments 2

Maya K
finally some real money going into defensive tools instead of chasing threats after the fact
tommy85
$1B valuation already? that’s bold but maybe they’ve got the models to back it up

Leave a comment