Tel Aviv-based cybersecurity startup Corma has raised $60 million in seed funding to develop artificial intelligence for defensive cybersecurity. The round was led by Sequoia Capital with participation from Khosla Ventures and Coatue.
The company, founded in 2025, works with Fortune 100 and Fortune 500 organizations across Healthcare, financial services, energy, critical infrastructure, and retail. Corma is building a foundation model specifically for defensive cybersecurity that powers autonomous AI agents capable of analyzing security data, correlating signals, and responding to complex attack campaigns. The new capital will be used to expand its interdisciplinary team, invest across pre-training, post-training, and cybersecurity research, and scale its AI-powered defensive workforce for enterprise customers.
The funding comes as artificial intelligence reshapes both cyberattacks and security operations. General-purpose AI models have become increasingly capable of coding, reasoning, and executing complex multi-step tasks. However, defensive cybersecurity presents a different challenge: security teams must process enormous volumes of logs, network activity, and alerts while connecting weak signals across long periods and maintaining accuracy across thousands of decisions. Corma’s research highlights this imbalance. In simulations involving enterprise environments, AI attackers succeeded in 88 percent of cases, while the same models detected only 12 percent of the threats they had previously created. These capabilities also support offensive cybersecurity activities, enabling attackers to identify vulnerabilities, develop exploits, and coordinate sophisticated attack campaigns. The race to general intelligence in cybersecurity has already begun, and attackers have a significant head start.
Corma is developing foundation models that can power specialized agents across defensive security functions. The company says these agents can operate across different security tasks while continuously learning the environments around them. Early deployments have shown significant operational improvements: Corma says its AI workforce has reduced threat response times by more than 94 percent and expanded security coverage by 15 times across different functions. The team includes AI researchers with experience at Google and DeepMind alongside experts from Israel’s 8200 Unit and major cybersecurity organizations. Shaun Maguire, a partner at Sequoia Capital, said Corma had trained its model for the complexity of real-world attacks and that the company’s technology provides the intelligence layer required for modern cyber defense.
The investment reflects growing concern about the speed of AI-fueled attacks. Businesses across critical services are under increasing pressure to intercept and deal with threats before they damage operations. Vinod Khosla, founder of Khosla Ventures, noted that critical services, health, and critical infrastructure could be impacted by automated cyber-attacks. Corma posits that AI will be crucial in closing the offensive-defensive gap, providing security staff with faster analysis, extended coverage, constant operational assistance, and improved threat detection capabilities. With continued expansion of AI by enterprises, Corma envisions a cyber defense built upon defensive AI and will use the new funding to develop the technology further and deploy it across high-value sectors.
