Corma, an Israeli startup building autonomous AI agents for cybersecurity defense, has raised $125 million in Series A funding at a $1 billion valuation. NEA led the round, with Sequoia Capital and Khosla Ventures also participating, the company announced November 16, 2026. The raise comes roughly three months after Corma closed a $60 million seed round, pushing its disclosed total funding to $185 million.
What does Corma's AI actually defend?
Corma builds proprietary foundation models, rather than relying on third-party large language models, designed to act as autonomous agents patrolling an organization's digital perimeter. The models are trained to find anomalies buried in enterprise telemetry, logs and events, the kind of \"needle in a haystack\" signal that indicates an intrusion or software vulnerability before it becomes a breach.
Founder and CEO Alon Pluda, a former Israeli military intelligence (Unit 8200) officer who previously co-founded Miggo Security, built Corma around the bet that general-purpose AI models are too broad for security-specific pattern recognition. The company, founded in 2025, now has 20 employees split between Tel Aviv and San Francisco, and counts Fortune 100 and Fortune 500 clients across healthcare, financial services, energy, critical infrastructure, retail and telecommunications. Corma says it is generating tens of millions of dollars in annual recurring revenue less than two years after founding.
Why raise again just three months after the seed round?
Demand outran infrastructure. Corma says client growth forced it to expand server capacity 24-fold to keep its defensive agents running at the scale its new enterprise customers require. That kind of compute scramble, rather than a product pivot, is the stated reason for stacking a $125 million Series A so soon after a $60 million seed.
Who's backing Corma, and how does the valuation stack up?
NEA led the Series A; Sequoia Capital and Khosla Ventures both returned or joined as participants. A $1 billion valuation on a company with reported tens of millions in ARR and 20 employees is a steep multiple, reflecting investor appetite for autonomous security agents broadly rather than Corma's revenue alone. The seed-to-Series-A gap of three months, with a valuation jump to unicorn status, signals that Pluda's pitch, specialized models beating general-purpose ones on narrow, high-stakes detection tasks, found quick buyers among top-tier funds.
How does Corma fit into the broader AI-agent security market?
Autonomous AI agents for security operations have become one of the most heavily funded corners of enterprise AI, as companies look to replace or augment human security analysts who can't keep pace with telemetry volume. Corma's differentiation claim, that purpose-built models outperform wrapped general models like GPT or Claude on security-specific anomaly detection, is the same argument driving a wave of vertical AI startups across other industries. Whether that edge holds as foundation model providers improve at narrow tasks is the open question for Corma and its peers.
What comes next for Corma?
The immediate priority is converting the fresh capital into more compute and headcount to serve the Fortune 500 pipeline the company says it already has in hand. Investors will be watching whether Corma's ARR climbs fast enough to justify the $1 billion mark before the next funding conversation, and whether its specialized-model thesis holds up against bigger AI labs moving into security.
