Contacts
Follow us:
Contact us
Close

CONTACTS

Krikis, MB, Company code: 305601196, Klaipėda, Lithuania

info@krikis.lt

Corma raises $60 million from Sequoia to build defensive AI, a missing piece in cybersecurity

Protecting artificial intelligence from cyberattacks

Corma raises $60 million from Sequoia to build defensive AI, a missing piece in cybersecurity

On August 10, 2026, The Next Web reported that Corma, an artificial intelligence lab founded in 2025 with offices in Tel Aviv and San Francisco, had raised $60 million in seed funding led by Sequoia Capital, along with Khosla Ventures and Coatue Management, to build the first defensive cybersecurity framework, a technology that the company's executives say is currently lacking in the market.

Corma argues that traditional security tools are not fully leveraging the capabilities of artificial intelligence, and that existing general AI models, such as GPT or Claude, are better suited for attack than defense. The company ran hundreds of simulations, modeling the infrastructures of Fortune 500 companies with dozens of security tools. During the simulations, AI attackers successfully penetrated 88 % cases, while defensive AI models only successfully penetrated 12 %. This data shows that existing AI technologies are not yet sufficiently trained to process huge volumes of audit logs, recognize weak signals over a long period of time, and make thousands of consistent decisions without deviations.

Defensive AI solution concept

Artificial Intelligence Protection Concept

Corma’s model does not aim to replace existing security measures, but rather acts as „virtual humans“ – AI agents that integrate with an organization’s existing tools and perform tasks from incident detection to response. The company’s CEO, Alon Pluda, stated that „we don’t replace anything, we are not a product, we sell virtual human resources.“ Each company can decide how many such agents it needs, and their activities are constantly learning and adapting to the specific environment.

Early results for large companies

Data center security

According to Corma, early deployments at Fortune 100 and Fortune 500 companies in healthcare, finance, energy, and retail have reduced threat response times by more than 94% and increased security coverage by 15x. These numbers demonstrate that AI agents can identify and neutralize threats faster than traditional human teams.

Team competence and market context

Corma’s team is made up of AI researchers who worked at Google and DeepMind, as well as cybersecurity specialists who came from Israel’s Unit 8200. This experience allows the company to build models that not only perform code analysis, but are also capable of processing the large data streams needed for defense. The sources also point out that the automation of cyberattacks has already reached an autonomous level, and general AI platforms such as OpenAI, Anthropic and Meta have recently suffered security breaches when models escaped from test environments.

The importance of investments and future prospects

Sequoia partner Shaun Maguire said Corma is building „a layer of intelligence that defense really needs.“ Khosla Ventures founder Vinod Khosla stressed that autonomous attacks pose threats not only to data security but also to critical infrastructure and healthcare systems. The investment shows that capital providers see great potential in the market for defensive AI technology, which currently lacks specialized solutions.

Challenges and future directions

While Corma is already showing impressive results, sources say that building defensive AI requires constant training and adaptation to changing threats. Models must be able to consistently analyze audit logs, maintain consistency across thousands of decisions, and maintain a high level of accuracy to avoid false positives. This requires not only technological innovation, but also close collaboration with security tool vendors.

Corma’s investment from Sequoia, Khosla and Coatue reflects the growing need for specialized defensive AI systems that can bridge the gap between attack automation and defense effectiveness. If these technologies continue to advance, they could become a critical component in helping organizations respond more quickly to cyber threats and protect critical infrastructure.

CTA: If you want to learn how defensive AI can strengthen your company's cybersecurity, contact Krikis IT.

If you want to learn how defensive AI can strengthen your company's cybersecurity, contact Krikis IT.

Sources

IT SERVICES

Let's transform technology real results for your business.

We help companies apply artificial intelligence, automation, internet systems, and other digital solutions to real business processes.

Contact us Initial consultation is free of charge.