Noma is building tools to spot security issues with AI apps

Companies increasingly recognize the vulnerability of their cybersecurity as they rapidly adopt artificial intelligence (AI), according to a recent survey of over 350 IT leaders. The findings reveal that more than half of the executives believe the complexity of AI applications has weakened their organizations’ cybersecurity posture, while over two-fifths express concern that their security teams lack the requisite skills to safeguard AI systems and workloads.

In response to the heightened demand for securing AI applications, a new wave of startups has emerged. Companies like HiddenLayer and Protect AI specialize in defending AI systems against adversarial attacks, whereas Cranium provides visibility into AI applications at the application layer. Among these newcomers is Noma Security, a startup co-founded by Niv Braun and Alon Tron, which aims to address comprehensive security challenges within the AI data lifecycle.

Noma Security, which recently emerged from stealth, seeks to develop tools designed to pinpoint vulnerabilities within data pipelines and code in data science environments. The startup targets various threats, including prompt injection attacks. Braun’s motivation for founding Noma stemmed from his experience at Verint, where he observed significant gaps in security practices concerning the unique intricacies of the data and AI lifecycle.

Braun highlights that existing security tools primarily cater to traditional software lifecycles, neglecting the distinctive R&D processes and technical vulnerabilities associated with AI models. Noma is positioned to rectify this by identifying and remedying misconfigurations in AI application components, such as recognizing sensitive data in training datasets that may inadvertently include personally identifiable information.

Noma’s unique proposition lies in its ability to centralize AI application security functions into a singular dashboard operable in both cloud-based and self-hosted environments. Braun articulates that the interface will empower companies to understand their security landscapes better, aiding in governance and the minimization of risks. This streamlined approach is intended to bolster confidence, enabling organizations to amplify their AI usage in more critical business domains.

Founded in 2023, Noma has already secured paying customers, including notable Fortune 500 firms in sectors such as B2B software, financial services, and retail. The startup recently announced a successful $25 million Series A funding round, which Braun noted was proactively pursued by investors despite having existing capital.

The latest funding round was led by Ballistic Ventures, increasing Noma’s total capital raised to $32 million, which also includes a previously undisclosed $7 million seed round initiated by Glilot Capital Partners and supported by Cyber Club London. Braun revealed plans for growth, stating that Noma aims to double its existing team of 20 within the next year to bolster development and market presence.