AI security startup Noma Security closes $100M funding round – News

Noma Security Inc., a developer of artificial intelligence security tools, has raised $100 million in funding to accelerate its engineering efforts.

Evolution Equity Partners led the Series B investment with participation from Ballistic Ventures and Glilot Capital. The round’s announcement today comes less than a year after Noma’s launch. According to the company, its customer base already includes several dozen organizations.

“AI agent adoption is exploding within our customer base and CISOs understand that AI innovation must be thoughtfully deployed with full guardrails,” said Noma co-founder and Chief Executive Officer Niv Braun (pictured, right, with co-founder and Chief Technology Officer Alon Tron).

Workers sometimes provision AI applications without the information team’s knowledge. As a result, administrators can overlook those applications during cybersecurity assessments, leaving potential vulnerabilities unnoticed. Noma has built a software platform that addresses such visibility gaps by automatically mapping out all the AI applications in a corporate network.

The software visualizes the assets it finds in a dashboard. Noma displays the employee responsible for each workload, as well as any vulnerabilities the workload may contain.

The company says its platform can find poisoned datasets, or AI training datasets into which hackers inject malicious content. Such content is usually meant to change AI models’ behavior. For example, hackers might modify a virus detection algorithm’s training dataset in a way that causes it to overlook certain malware strands.

Noma also spots a range of other cybersecurity problems. It can detect weak points in the infrastructure that powers an AI application, vulnerable open-source components and malicious models. The software organizes the issues it finds by severity to help administrators prioritize their remediation efforts.

Another set of features in Noma’s platform focuses on AI agents. The software maps out the systems to which an AI agent has access and scans them for vulnerabilities. It also spots related issues, such as configuration settings that give an agent access to more data than it strictly requires.

The platform collects security data on AI workloads from the development tools that companies use to build them. As a result, it can spot vulnerabilities before they roll out to production. After the rollout, the platform provides guardrails that filter malicious prompts.

Securing AI applications is not the only task that Noma promises to speed up for developers. According to the company, its platform can also check how well workloads adhere to regulations such as the EU AI Act. While at it, the software checks evaluates whether applications comply with the licensing terms of the open-source components they incorporate.

Today’s funding round will enable Noma to grow its engineering team and speed up product development initiatives. In parallel, the company plans to scale up its international go-to-market efforts. 

Photo: Noma Security