Case Study

From shadow AI to governed adoption in two weeks

An anonymized look at a 40-person consulting firm that took two weeks to move from 'we know our team is using ChatGPT and Claude but we haven't talked about it' to a written policy, trained staff, and tracked adoption.

Reviewed by Level Up Automate.
TL;DR
  • Inventory revealed 11 distinct AI tools in use across 40 staff before any policy existed.

  • Two weeks: inventory → policy draft → leadership review → all-hands rollout.

  • Six months later: 32 of 40 staff using approved tools regularly; zero reported incidents.

Starting point

The firm had quietly become reliant on ChatGPT and Anthropic's Claude for first-pass client deliverables and Otter for meeting transcripts. Leadership had not formally approved any of them. The CEO booked a 30-minute call after a peer firm published an embarrassing AI-related news item.

What we did

Two weeks, four steps.

  • Day 1–3: anonymous tool inventory across the team.
  • Day 4–7: one-page policy drafted, leadership review, edits.
  • Day 8–10: 60-minute all-hands rollout, signed acknowledgments.
  • Day 11–14: vendor due diligence on the two highest-priority tools.

Outcomes at six months

Active use of the approved primary tools (ChatGPT Team for general drafting; Anthropic's Claude for longer client analysis) across 32 of 40 staff. Two near-misses self-reported and resolved before any client impact. A new enterprise prospect cited the firm's clear AI posture as a tiebreaker in their decision.

Common questions

Plain-English answers

Was anyone resistant?
Two senior staff were skeptical — both wound up among the heaviest users within three months once the rules were clear.
What did it cost?
Engagement fee plus tooling. Less than the cost of a single botched proposal.
Next step

Want a hand getting this right?

A 30-minute conversation often saves weeks of guessing. We'll talk through your team, your data, and what to do first — no slide deck required.