NordLayer AI Scout Aims to Tame the "AI Zoo" in Corporate Offices

NordLayer has introduced NordLayer AI Scout, a new platform designed to help companies manage the proliferation of artificial intelligence tools, which the company dubs an "AI zoo." The tool aims to provide IT departments with a clear view of both approved and unauthorized AI applications being used by employees across web and desktop environments.

According to Andrius Buinovskis, VP of product strategy at NordLayer, organizations are facing a growing number of unsanctioned large language models (LLMs) in daily use, increasing the risk of data leaks and security breaches. Employees often adopt these tools for convenience, such as using a preferred chatbot for summarization or a familiar coding assistant, or due to lengthy approval processes for company-sanctioned software.

Research cited by NordLayer indicates that 70% of employees utilize free public LLMs at work, compared to 42% who use company-provided tools. This highlights a gap between formal IT policies and actual employee behavior, where workers may independently develop AI usage habits.

NordLayer AI Scout monitors on-device activity to measure actual tool usage, rather than relying solely on software license counts. This allows companies to identify which AI tools are gaining traction, which licenses are underutilized, and the cumulative token spend across deployed LLMs.

The platform also incorporates data loss prevention (DLP) features focused on user prompts. It can inspect content before submission, applying company policies to warn users, redact sensitive information, or block submissions entirely. Vakaris Noreika, VP of product research and development at NordLayer, stated that the AI DLP is built to inspect and control prompts pre-submission to prevent sensitive data from being inputted.

This approach offers an alternative to outright bans on unfamiliar AI tools, enabling security teams to observe where AI activity is occurring and identify behaviors that genuinely require oversight. The tool is also designed to catalog connections made by AI agents through Model Context Protocol servers, inventorying the capabilities of each agent and what external systems they can access.

NordLayer suggests that the rapid emergence of new AI tools and the speed at which employees adopt them make traditional software approval models increasingly difficult to maintain. AI agents can also acquire new connections after initial approval, complicating IT management. NordLayer AI Scout is positioned not as a tool to block all unapproved AI, but rather as an inventory management and control system for AI activities that pose a significant risk.