---
title: AI tool adoption
canonical: "https://cloudmonitor.ai/docs/using-cloudmonitor/reports/ai-adoption/"
description: "Which AI tools your teams actually use, how spend splits across the business, who the power users are, and which seats are idle."
---

:::note[Not everyone has this]
**Tokenomics** is an optional module. If you can't see it in the menu, it isn't switched on for
your workspace — ask an admin to turn it on under Settings ▸ Workspace Settings ▸ Features.
:::

Four screens answering the question a finance director actually asks about AI: *are we getting
anything for this?*

They're separate menu items, and they're best read together.

## AI Tool Adoption

> Which AI tools your teams actually use, ranked by active users.

The starting point. Organisations typically end up paying for more AI tools than anyone realises,
because they arrive team by team. This is the inventory with usage attached.

Two tools doing the same job, both half-used, is the pattern worth looking for.

## AI Spend by Business Unit

> How your AI and machine-learning spend splits across your cost groups.

The chargeback view. It uses the same cost groups as the rest of CloudMonitor, so AI spend lands in
the same business hierarchy as your cloud spend rather than in a category of its own.

## Power Users

The people using AI tools most, with their usage and spend for the period.

Useful in two directions, and the second is the one people miss:

- **Cost control** — is a small number of people driving most of the bill?
- **Enablement** — what are the heaviest users doing that everyone else isn't? A power user is
  usually a training opportunity rather than a problem.

## Seat Utilization

Seats you're paying for against seats anyone is actually using, plus the idle-seat gap.

This is the clearest money on these four screens. A per-seat AI licence assigned to someone who
hasn't used it in two months costs the same as one in daily use, and reclaiming it needs no technical
change at all.

## What's live

:::tip[Per-user data is GitHub Copilot only, on live data]
Copilot is seat-licensed and its per-user feed is connected. The other tools — Anthropic, OpenAI,
Cursor — appear in adoption and spend but their per-user detail is not wired yet.

Filter Power Users or Seat Utilization to a non-Copilot provider on live data and you'll get an
honest empty state, or "not seat-licensed" — never invented per-person figures.
:::

## Filters

Period, provider and cost basis apply. On AI Tool Adoption the provider chip filters by tool, which
is the useful thing to do here.

Cost group and virtual tags are greyed out on the per-user screens: per-user AI usage isn't allocated
to cost groups yet, and it carries no resource tags. **AI Spend by Business Unit is the exception** —
that one is scoped by cost group, which is the whole point of it.

## Related

Cost detail is on [AI Spend](/docs/using-cloudmonitor/reports/ai-spend/); efficiency suggestions are
on [TokenMaxing](/docs/using-cloudmonitor/reports/tokenmaxing/).

Per-person AI **budgets** can be set today, but the spend against them is an example allocation until
those per-seat feeds land — see
[what isn't available yet](/docs/reference/whats-not-available-yet/).