Three components. One governance layer.
Brolli AI is a browser extension and a desktop agent for your employees, and a dashboard for your administrators. Together, they give your organisation real visibility and real control over AI tool usage.
Sits in the background. Steps in when it matters.
The Brolli AI browser extension is installed on employee devices through your standard device management process. Once it's there, it runs quietly in the background.
Brolli AI is built on Chromium, which means it works across any Chromium-based browser, including Google Chrome, Microsoft Edge, Brave, and Opera.
When an employee visits an AI tool site, the extension checks it against your organisation's policy in real time. Depending on how you've classified that tool, it either shows a brief information badge, asks for an acknowledgement, or prevents access.
- A weekly summary of their AI tool activity
- The current acceptable use policy for tools they're using
- A clear explanation whenever a tool requires acknowledgement or is blocked
- Their own activity log (not shared across the organisation)
Their colleagues' activity. Individual usage is not surfaced to other employees or to line managers. Brolli AI is an organisational tool, not a performance monitoring system.
The Brolli Stealth desktop agent catches what the extension can't.
A lightweight background agent for macOS, Windows, and Linux, deployed the same way as the extension — through your device management system. It detects native AI desktop apps (ChatGPT, Claude, Cursor, GitHub Copilot in your IDE) and AI tool usage in any browser the extension isn't installed in.
It runs in the background — “stealth” refers to how little it asks of your team, not to how visible it is. A tray icon is always present so employees can see it's running, mirroring the same colour tiers as the extension.
Linux: manual setup only for now — enter your API key directly in the tray popover. Managed deployment (autostart, MDM-pushed config) isn't built yet, unlike macOS and Windows.
Needs an API key from your organisation's Brolli AI account to activate. Book a walkthrough if you're getting set up.
- Native desktop AI apps not installed as browser tabs
- AI tools used in non-Chromium browsers
- AI coding assistants running inside your IDE
Self-install or push it through your device management system — either way, your IT or compliance team should share our employee-facing explainer before rollout, so nobody is surprised by the tray icon showing up.
A policy response that matches the actual risk.
Every AI tool in our dataset has a risk classification. You can accept our defaults or customise them for your organisation. The classification determines how Brolli AI responds when an employee visits that tool's site.
The employee sees a brief badge confirming the tool is approved and showing basic information: tool name, category, data residency. No interruption to their workflow.
Use this for: ChatGPT Enterprise, Microsoft Copilot, Google Workspace AI features, and other tools your organisation has formally approved.
Before the employee can access the tool, they're asked a short question: what are you planning to use this tool for? Their response is logged. The acknowledgement is recorded.
Use this for: tools you haven't blocked but want visibility on, tools your organisation is evaluating, and any tool where understanding why employees are using it is as useful as knowing that they are.
Access is prevented. The employee sees a clear message explaining that the tool is not permitted under your organisation's policy, with a link to your acceptable use policy. No ambiguity.
Use this for: tools known to train on user data, tools with no enterprise privacy controls, tools that fall outside your data sovereignty requirements.
Visibility from day one.
The Brolli AI admin dashboard gives your IT administrators, compliance officers, and leadership a clear view of AI tool usage across the organisation.
The AI tool landscape changes constantly. Our library changes with it.
Most governance tools ship with a static list of domains to block. Brolli AI is different. The tool dataset at the heart of the product is actively maintained, continuously expanded, and automatically re-scanned.
When a tool changes its terms of service, gets acquired, or shifts how it handles user data, the risk classification updates automatically. Your governance coverage expands as the market expands.
A tool classified as Caution has its terms updated. Model training on user inputs is enabled by default. Brolli AI detects the change, reassesses the tool across all six dimensions, and reclassifies it as Blocked. Your policy updates automatically. Employees who next visit that tool see a block nudge and a clear explanation. You didn't have to do anything.
Up and running in under a day.
Brolli AI is designed for organisations that don't have a dedicated security engineering team. If you have a Microsoft 365 or Google Workspace environment, you already have what you need.