NEA led Corma’s $125 million Series A as the cyber defense AI startup said it has raised $185 million total.

Corma raises $125M Series A at $1B valuation

Corma raises $125M Series A at $1B valuation

Image: AI-generated editorial illustration

Corma announced a $125 million Series A on November 16, led by New Enterprise Associates, valuing the defensive cybersecurity AI startup at $1 billion. The Tel Aviv- and San Francisco-based company said the new capital follows its August seed round and stealth exit, bringing its reported total funding to $185 million.

Corma said on November 16 that it raised a $125 million Series A led by New Enterprise Associates at a reported $1 billion valuation, a rapid follow-on to the $60 million seed round it announced roughly three months earlier when it emerged from stealth. Secondary reporting from TechCrunch and Israel Defense said the round brings Corma’s total reported funding to $185 million.

The company builds specialized AI foundation models for defensive cybersecurity and deploys them through autonomous agents that work alongside enterprise security teams. Corma’s system is designed to analyze audit logs, security events and network traffic through customers’ existing security tools, with the goal of correlating weak signals across environments and helping teams respond to attacks. SecurityWeek reported that Corma emerged from stealth in August with a frontier foundation model for defensive cybersecurity, while the company’s backers have described it as an autonomous AI security workforce.

Who backed Corma's Series A?

NEA led the round, and existing investors Sequoia Capital, Khosla Ventures and Coatue reportedly participated. NEA publicly confirmed on November 16 that it was announcing a partnership with Corma and said it was leading the Series A.

In its announcement, NEA described Corma as "the first frontier AI lab for defensive cybersecurity" and said the startup is building a superintelligent foundation model trained natively on defensive security data. NEA also said Corma’s workforce deploys within minutes, works through customers’ existing collaboration and security tools such as Slack, Teams and email, and can request approvals when needed while autonomously executing complex tasks end to end.

What does Corma's product do?

Corma’s product is meant to help security teams handle large volumes of operational data, including audit logs, network flows and alerts. According to NEA and other reports, the system is built to correlate faint signals across tools and over time, rather than treating security data as generic text or code.

NEA said Corma trains its model with adversarial methods and large-scale reinforcement learning in realistic enterprise environments. It also said the model outperforms general-purpose large language models on quality, speed and token cost, though those performance claims were attributed to the company and its investor rather than independently benchmarked in the reporting.

The company says its agents can be deployed inside existing security operations workflows. SecurityWeek reported that the system is intended to audit security data and correlate signals across environments to improve defenses and help teams respond to cyberattacks. Pulse 2.0 said Corma deploys its models through an autonomous AI workforce operating inside a customer’s existing security tools.

Why is the startup drawing attention so quickly?

Corma was founded in 2025 and is headquartered in Tel Aviv and San Francisco, according to SecurityWeek. Israel Defense reported that the company says demand has surged among Fortune 100, Fortune 500 and Fortune Global 500 companies since it emerged from stealth, and that it expanded server capacity 24-fold after its launch.

NEA said Corma already had multiple production deployments across Fortune 500 and other large enterprises three months after its stealth exit. The firm also said the startup’s AI security workforce has already autonomously detected, investigated and contained active cyberattacks in real enterprise environments, and that customers are expanding its responsibilities from threat hunting and incident response into detection engineering and digital forensics.

Pulse 2.0 reported that Corma has not disclosed customer counts, revenue or detailed deployment metrics. The same report said the Series A is expected to support expansion of research, computing capacity and deployments with large organizations, though no itemized use-of-proceeds breakdown was provided.

Israel Defense also reported that Corma’s team includes researchers from Google and DeepMind as well as cybersecurity specialists from Israel’s 8200 intelligence unit. NEA similarly said the company’s team includes researchers from DeepMind, Microsoft and other frontier labs, along with hackers from industry. Corma’s rapid fundraising, high reported valuation and early enterprise traction place it among the more closely watched defensive AI startups in cybersecurity.

Sources