Browser
AI chat products opened in a browser tab.
See every AI tool your organization actually uses
Map every model and surface employees use, sanctioned or not, through metadata-only telemetry, so shadow AI becomes visible and governable.
Example: Claude, ChatGPT, Gemini and Copilot are sanctioned, Meta AI and Qwen are blocked. DeepSeek is detected as a new tool, marked Unreviewed, and an admin sets it to Unsanctioned.
Why it matters
New AI tools arrive without anyone asking IT.
People sign up for whatever helps them work faster, often with a personal account. The tool never goes through vendor review and never shows up in SSO logs.
What it finds
Every AI tool, on every surface
Alesia detects AI providers from what actually runs on the device. There is no fixed list of known products, so a brand-new tool shows up the same way an established one does.
IDE
Plugins and assistants that call a model from the editor.
CLI
Terminal sessions that reach a model with an API key.
Desktop apps
AI apps installed on the device.
Request volume per tool over timeNo surveyNo manual scan
Decisions
Sanction, review, or block
Every detected tool gets a status. Shadow AI turns into a list of decisions instead of a standing worry.
Sanctioned
Approved for use. Add it to your inventory, where its usage and cost are tracked from then on.
Unsanctioned
Not approved. Its requests count toward the unsanctioned share in your overview.
Needs review
Flagged for a closer look before you decide.
Blocked
Requests are blocked on the devices your policy covers. Each attempt is recorded as an incident.
Risk
Start with the tools that matter most
Every detected tool carries a risk level, from low to critical. The overview shows how many tools are in use, how many are critical, and what share of all requests goes to tools you haven't sanctioned, so the review starts at the top of the list.
Metadata only
Visibility without surveillance
Alesia surfaces which AI tools are in use, by whom, and how often, without reading the content of a single conversation. IT and security get a real inventory instead of a guess.
Use cases
See it in action
Shadow AI, in detail
No. Alesia detects AI providers from what actually runs on the device, so a brand-new tool shows up the same way an established one does.
No. Shadow AI is built on metadata: which tool, which runtime, how many requests. Prompts and responses never leave the device.
Requests to that tool are blocked on the devices your policy covers. The attempt is recorded as an incident, so you can see whether people still reach for it.
Say yes to AI.
Set the rules once. Alesia enforces them on every device, and prompts stay on your devices.