Reco launches Browser Guard to curb shadow AI risks

Reco has launched Browser Guard, a browser-based security product for employee use of artificial intelligence tools. The product is now generally available.

The launch addresses a growing concern for security teams as workers use AI assistants, coding tools and automated agents through web browsers rather than within software environments already under central oversight. Browser Guard is designed to monitor prompts and agent activity in the browser and apply policy controls before sensitive data is exposed or an unauthorised action takes place.

For many organisations, the issue is the spread of AI use beyond approved applications. Staff may interact with services such as Claude, Copilot and Gemini through corporate or personal accounts, while browser extensions and third-party tools add another layer of risk. Reco said its product is intended to show which AI tools are being used, who is using them, whether the account is personal or corporate, and whether business information is being shared.

Reco is positioning the product as a control layer at the point where AI activity begins. Most employee interaction with AI tools starts in the browser, according to the company, so organisations need to inspect not only the text entered into prompts but also the identity of the user, the type of data involved, the destination of that data and the business purpose behind the interaction.

That focus reflects a broader shift in security as AI systems move beyond simple question-and-answer tools. Agents can retrieve context, call external tools and take actions on behalf of users, extending risk well beyond the original prompt, Reco said. In that model, a seemingly ordinary query may become more significant once it is tied to customer data, internal systems or automated workflows.

Shadow AI

A central element of the launch is visibility into what the industry often calls shadow AI, where employees use unapproved tools outside official company channels. Browser Guard can identify AI services, browser extensions and personal accounts that fall outside approved use, giving security teams a way to spot activity that might otherwise sit beyond their monitoring systems, according to Reco.

That matters because security teams are increasingly being asked to balance two pressures at once: allowing staff to use fast-changing AI tools in day-to-day work while limiting the risk of confidential information being entered into external systems or automated actions being carried out without the right checks. Browser Guard is meant to let teams detect, guide, block or stop activity according to internal policy.

Reco also said the product evaluates actions taken by agents rather than looking only at prompts typed by employees. The distinction matters because an agent may go on to access files, pull in extra context or trigger a workflow after the initial instruction has been submitted. In practice, that makes the security question less about a single message and more about the chain of actions that follows.

The company has positioned itself around what it calls agent ecosystem security, focusing on the links between applications, identities, permissions and workflows. Browser Guard fits that approach by adding controls in the browser while drawing on wider context from the rest of the organisation's software environment, Reco said.

Reco said its wider platform includes more than 280 application integrations and 1,000 detection controls. Those figures point to a strategy centred on broad visibility across third-party software and user activity, especially as security teams face a more fragmented landscape of AI services, cloud applications and automated agents.

Ofer Klein, Co-Founder and Chief Executive Officer at Reco, outlined the company's argument for the new product.

"AI security is not just about knowing which tools employees use," said Ofer Klein, Co-Founder and Chief Executive Officer at Reco. "Security teams need to know what data enters each interaction, what an agent can reach and whether its next action should be allowed. Browser Guard enforces policy decisions in the browser, where AI use is actually happening, with the context needed to stop risky activity before it creates exposure."

The launch comes as security vendors race to address risks created by widespread employee use of AI tools, especially through everyday browser sessions that can fall outside traditional monitoring systems. Reco's position is that the browser is now the main point of control for that activity because it is where prompts are entered, accounts are chosen and automated tasks are initiated.

For companies trying to manage AI use without banning it outright, the product is aimed at a practical operational problem: how to distinguish acceptable use from risky behaviour in real time. Browser Guard is available now.