soft-shell crab exportersoftshell crab exporterVietnam crab exporterVietnamese mud crab export
🏈 Draft. Trade. Win. Black Friday 2026 🛍️ Check home prices 🏠 Discover PLAY 🤩

What Employees Paste Into AI Might Be a Security Problem

Adobe Stock
K.H. Koehler
Contributor
Oct. 16, 2026, 5:58 p.m. ET

An employee does not have to upload a confidential file to expose company information through AI. Pasting a few lines into a chatbot can be enough.

NordLayer is launching NordLayer AI Scout around that problem. The new AI visibility and security solution is meant to monitor AI use across web and desktop and can inspect prompts before they are sent, which aims to give companies a chance to intervene when sensitive material is about to leave their control.

The product also shows which AI tools employees are actually using, including services that were never approved by IT. NordLayer cites research showing that 70% of workers use free public LLMs on the job, compared with 42% who use company-provided tools, leaving many employers with a wider AI footprint than their software inventory suggests.

NordLayer AI Scout’s built-in data loss prevention is designed specifically for AI. Depending on company policy, it can warn an employee about a prompt or redact sensitive information before submission. It can also block the prompt when the risk is high enough.

“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 timing can be important because traditional data-loss systems were built largely around files and known routes out of a company. Generative AI may give employees another way to move information outside those boundaries without thinking of the action as a transfer at all.

A developer might paste a block of proprietary code into a tool for help troubleshooting it. Someone in sales could drop part of a customer conversation into an LLM to clean up an email. The employee may be trying to save a few minutes rather than bypass security, which makes the prompt itself a more useful place to intervene than a warning issued after the information has already been sent.

IT Can’t Govern Tools It Doesn’t Know About

NordLayer AI Scout also addresses the problem of shadow AI, the growing collection of tools employees adopt without formal approval.

“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 AI Scout is designed to monitor activity on-device and to bring approved and unauthorized use into one view. Security teams can see which tools have taken hold across the workforce and decide where controls belong instead of banning unfamiliar services simply because they lack visibility.

The platform also measures actual adoption rather than relying on purchased seats. NordLayer says companies can compare active use with paid licenses and estimate token spend across deployed LLMs, which can expose another mismatch: a business may be paying heavily for one platform while employees have quietly settled on something else.

NordLayer cites IBM research showing that breaches involving shadow AI carry about $400,000 in additional cost above the reported global average breach cost of $4.99 million. Unauthorized AI use can be especially difficult to contain when it sits outside the monitoring already applied to approved software.

NordLayer AI Scout also watches AI agents, which create a different type of exposure because they can connect directly to other systems. The platform is designed to maintain a live catalog of Model Context Protocol servers used by coding agents and record the skills available to those agents. This is meant to give IT teams a clearer view of what each one can reach after it has been connected.

View the NordLayer AI Scout demonstration here.

The release of NordLayer AI follows the company’s launch of NordLayer Browser earlier this year and aims to push its security products deeper into workplace AI use.

Employees are already experimenting with AI because it helps them get work done. NordLayer’s launch is aimed at the part companies have struggled to see: what people are typing into those tools, and where the information goes once they hit send.

More from Contributor Content  

Our editors independently choose our recommendations. Some content is produced with paid support from a third party, however our editorial decisions remain independent. If you buy through our links, the USA TODAY Network may earn a commission. Prices and availability may change.