Cybersecurity firm NordLayer has launched a new product called NordLayer AI Scout, designed to help employers track the growing number of AI chatbots and coding assistants being used by their workforce, including tools that were never officially approved.
The company describes the problem as an “AI zoo” — a reference to the mix of sanctioned and unsanctioned AI tools that accumulate inside organizations as employees adopt whatever chatbot or assistant helps them get work done faster.
“What companies are facing today is an AI zoo — dozens of unsanctioned LLMs in daily use across the workforce, each one widening the organization’s exposure to data leaks and security breaches,” said Andrius Buinovskis, VP of product strategy at NordLayer.
NordLayer says this pattern is common and usually not malicious. An employee might prefer one chatbot because it writes better summaries, a developer might keep using a familiar coding assistant, or someone might turn to a free AI tool because getting access to the company-approved version takes too long. Individually, these choices rarely feel like rule-breaking, but collectively they can leave a company with a far larger AI footprint than anyone formally signed off on.
The company points to research showing that 70% of employees use free public large language models at work, compared with just 42% who use the tools their employer officially provides. That gap means a business can invest heavily in an approved AI platform while workers quietly build habits around something else entirely.
How the tool works
NordLayer AI Scout monitors AI activity on-device across web and desktop, giving IT teams a single view of which tools are actually being used rather than relying on licensing records alone. The platform tracks which tools have gained real traction among employees, which paid licenses are going underused, and how much is being spent on tokens across deployed AI systems.
The company says the stakes rise once employees start feeding company information into these tools. An employee does not need to upload a full document to create risk — pasting a few paragraphs of internal material or a block of proprietary code into a chatbot can expose that information just as easily, often without the employee recognizing it as a data transfer.
To address that, NordLayer AI Scout includes data-loss-prevention features built around the text employees type into AI tools before it is submitted. The system can inspect a prompt, apply company policy, and either warn the user, redact sensitive details, or block the submission outright.
“We built NordLayer AI Scout for how AI actually works — not just what it looks like from the outside, but for what’s happening at a technical level,” said Vakaris Noreika, VP of product research and development at NordLayer. “Its AI DLP is designed to inspect prompts and control them pre-submission, to ensure that sensitive data is never inputted in the first place.”
That approach gives employers an alternative to outright banning unfamiliar AI tools. Since many employees are already experimenting with them, security teams can instead identify where that activity is occurring and decide which specific behaviors need restrictions.
Tracking AI agents
The product also addresses a newer wrinkle: AI coding agents that connect to outside systems through what are known as Model Context Protocol, or MCP, servers. Because these agents can reach well beyond what’s visible on an employee’s screen, NordLayer AI Scout is designed to maintain a live catalog of those connections, giving IT departments insight into what each AI agent can access after it has been deployed.
NordLayer argues that this level of detail is increasingly necessary because traditional software approval processes, which vet one application at a time, are struggling to keep pace with how quickly new AI tools emerge, how fast employees adopt them, and how agents can gain new system connections after their initial approval.
Rather than positioning the product as a way to block every unapproved AI tool, NordLayer describes it as an inventory of what is already in use inside a company, paired with controls for the moments when that activity poses genuine risk.
You must be logged in to post a comment Login