See how your business uses AI. Set the rules. Keep the record.
Vireo Sentinel shows which AI tools your people use and what for. It puts your AI policy in front of them, and keeps the record that shows the rules are followed. ChatGPT, Claude, Gemini, Perplexity and Microsoft 365 Copilot, in one place.
"AI should be a competitive edge, not a risk to manage. Vireo lets us treat it that way."
It comes down to three things
Your people already use AI. You need to know what's used, set what's OK, and be able to show it.
See
Which AI tools your people use, how often, and what kind of work they use them for.
Set the rules
Put your AI policy in front of people, ask for a second look at risky prompts, and block the sites you don't allow.
Prove
A report and a record you can hand to a client, a board or an auditor.
Right now you're guessing which AI tools are in use
Vireo Sentinel shows you who is using which tools, and where the risk is.
- Prompt activity on ChatGPT, Claude, Gemini and Perplexity
- Visits to over 150 other AI tools
- Microsoft 365 Copilot activity in the same view
- AI at Work: what kind of work AI is being used for, and how that changes over time
Example: when AI gets used across the week
Example only
An AI policy nobody reads isn't a policy
Put the rules where the work happens, and get a record of who agreed to them.
Policy prompt
Your acceptable use policy shows when someone opens an AI tool. They agree before they can use it, and you keep the record.
A second look at risky prompts
In Intervention mode, a person whose prompt contains sensitive data can edit it, redact it, cancel, or send it anyway with a reason. In Insight mode the prompt is recorded and nobody is interrupted. How businesses use the two modes.
Site blocking
Block the AI tools you don't allow. People see your message and where to go instead.
When a client asks how you manage AI, hand them the record
The record is kept as people work, with nothing to compile.
- A governance report in six sections, citing the ISO/IEC 42001 clauses and Australian Privacy Principles they relate to
- Every review decision recorded, including the reason when someone sends anyway
- Controlled access to prompt text, with a history of who looked and when
What's in the governance report
- AI systems inventory
- Risk assessment and treatment
- Control effectiveness evidence
- Residual risk assessment
- Continuous improvement
- APP 1.7 automated-decision facts
Copilot and browser AI, in the same record
You pay for Copilot. Your people also use ChatGPT and Claude in the browser. Connect your Microsoft 365 tenant and both show up in one place. We read Copilot activity from your audit log. We don't collect Copilot prompt text.
See how Copilot worksSensitive data stays in your browser
Card numbers, bank details, tax file numbers, passwords and keys are removed before anything is sent.
The check runs in the browser
Over 90 patterns and keyword lists run on the person's own machine. Prompts aren't sent to an AI service to be checked.
Redacted before it's sent
A record of each prompt goes to Vireo Sentinel. Where sensitive data is found, it's redacted in the browser first, and again before it's stored. The original is never stored.
You decide who can read prompts
Prompt text stays hidden until an administrator confirms they're authorised to read it. Each period of access is logged in an access history.
See what AI is doing for the business
AI at Work sorts activity by the kind of work being done, so you can see where AI is earning its keep and where people need help. Trends shows how that moves over time.
Workflows picks out the tasks people repeat.
Example: what the AI is asked to do
| Explain or define | 24% |
| Draft or create | 21% |
| Edit or revise | 15% |
| Analyse or calculate | 12% |
| Review or check | 10% |
Installed as a browser extension
Nothing to change on your network.
Install
Push the extension to Chrome, Edge and Firefox with your device management. Each browser picks up its owner's licence.
Watch
Every organisation starts in Insight mode. Nobody is interrupted while you learn what's in use.
Set the rules
Turn on prompt review, the policy prompt and site blocking when you're ready.
Your clients are asking you about AI. Give them an answer
We're signing our first MSP partners. Run AI oversight for your clients and hand them the report.
Prices on the page
For teams of 5 to 1,000+. AUD, ex GST. A card starts the 14-day trial, and you're not charged if you cancel before it ends.
Starter
$55 /month
5 users included. $11 per extra user
Business
$220 /month
20 users included. $8 per extra user
Enterprise
$860 /month
100 users included. $5 per extra user
Common questions
No. Nobody at Vireo can read your prompt records unless you give us access to your account.
It's built to show how the business uses AI. Tell your team it's there: the policy prompt does that for you. Insight mode records without interrupting. Intervention mode shows people a prompt when something risky is about to be sent. Access to prompt text is limited and logged. Each person can also see their own activity from the extension icon.
No. People get a warning before they attach a file to an AI tool, and the file's name, type and size go on the record. The contents aren't scanned or collected.
Chrome, Edge and other Chromium-based browsers, plus Firefox.
Yes. Copilot activity comes from your Microsoft 365 audit log. See how Copilot works.
Know what's in use by the end of the week
Start the trial, install the extension, and see your first report.
14-day free trial. A card starts the trial, and you're not charged if you cancel before it ends.