Companies are concerned that their excessive enthusiasm for AI has made them vulnerable to cyberthreats. According to a recent survey of more than 350 IT leaders, over half of the surveyed executives said the sophistication of AI applications compromised their company's cybersecurity position. A separate survey found that two-fifths or better of executives believe that the security teams in their firms do not have the capability to protect their AI workloads and applications.

The need for solutions to secure AI apps created an entire class of startups. For example, HiddenLayer and Protect AI specialize in defending AI systems against adversarial attacks, while Cranium provides visibility into AI systems at an application level.

A new company emerging from stealth today, Noma Security does that and a lot more, says co-founder and CEO Niv Braun. The startup is building tools to detect vulnerable data pipelines and code in the data science environment and also threats like prompt injection attacks.

So, Braun left customer experience automation firm Verint, which was much on his mind. He found out a "significantly blind spot" for security teams in the data and AI lifecycle.
"What we have today is tools designed for the traditional software lifecycle, but the data and AI lifecycle is fundamentally different," said Braun. "It has different R&D processes, requires different technologies and has different technical vulnerabilities within AI models.".

Noma is designed to help find — and correct — misconfigurations in the components used to build and run AI applications. For example, the platform looks for "sensitive" data in model training datasets, say text with personally identifiable information.

As Braun has noted, companies are in a better position to prepare themselves to face such threats if they can consolidate AI app security functions in one dashboard that can be deployed in cloud-based and self-hosted environments.

"Many single-function security solutions are flooding the market in an attempt to capitalize on the AI hype," he said. "By enabling companies to understand their AI application footprint and providing governance controls, Noma is helping organisations not only minimize and mitigate risk but also providing them the confidence they need to expand AI usage faster and to even more high-value parts of the business."

Noma, founded in 2023, has already started receiving paying customers, Braun says-participants include Fortune 500 clients in B2B software, financial services, and retail. Just recently, the startup closed a $25 million Series A funding round that Braun said was preempted by investors; Noma had cash to spare.

"Taking the [$25 million] even with substantial funds left in the bank, we wanted to accelerate our development and particularly our go-to-market teams and efforts," he said, noting Noma will double its 20-member team in the next year.

Ballistic Ventures has led Noma's Series A round, which raises the company's total capital to $32 million. Noma has previously raised a seed round of $7 million that was not announced, led by Glilot Capital Partners, with participation from Cyber Club London.