AI Enablement · Case Study
Closing the AI Adoption Gap.
How a publicly funded organization closed the gap between AI power users and everyone else in a single half-day. No technical training. No new tools. Just judgment.
The situation
The licenses were bought. The return was uneven.
That judgment is teachable. It is also the part almost no one is teaching.
Most organizations we talk to are past the question of whether to give their people AI tools. The license is paid for. Copilot, Claude, ChatGPT, or some combination is already on every desktop. The real question is why the return looks so uneven.
A handful of people have run away with it. They have folded AI into how they write, analyze, and prepare their work, and they are visibly faster for it. Everyone else has tried it once or twice, got a mediocre result, and quietly went back to the old way of doing things. A few are steering clear altogether, because they have heard it makes things up.
The gap between those two groups is rarely about access or aptitude. Both groups have the same tool on the same desk. The difference is that one group has developed judgment about when to trust the output and when not to, and the other has not.
The approach
We focused on judgment, not features.
People do not struggle because they cannot find the right button. They struggle because they cannot tell a genuinely good output from a confident-sounding wrong one. The session was built around four moves.
01
Know the tool
What AI actually is, and what it is not, in plain language. No model architecture lectures.
02
Apply it
Where it genuinely helps the real work the team does every day, and where it does not.
03
Catch the errors
Spot confident, well-formatted, wrong output before you act on it. Practiced live on real examples.
04
Decide
A simple verification check every person can apply on their own, on any AI output, going forward.
What we ran
Real tasks. Live practice. A planted error.
We did not teach features. The session was built around real work the team already does and live practice running it through AI.
- Group exercises on actual tasks: drafting, summarizing long reports, sense-checking data.
- A planted-error scenario: a polished report that pinned 67% of a department's problems on one team. The 67% figure appeared nowhere in the source data. The group caught it on their own.
- Open debriefs after each exercise: what did it get right, what did it miss, what would you change before this went anywhere real.
What changed
A verification habit, not a tool feature.
- Every participant left with a specific task to use AI for in two weeks.
- Every participant left with a five-question check to run before acting on any AI output.
- The group caught the planted error independently. The exact failure mode that causes real damage in the wild.
- Staff who had avoided AI entirely left with a concrete, low-risk place to start.
Want this session for your team?
Half-day, tool-agnostic, judgment-first. Built for the people who are not yet power users.
In their words
Anonymized participant feedback.
"The considerations around using AI, I understood it is not the most reliable, but getting into the details of why and how helps shape how I strategically use it in the future."
"Loved the style. No pressure to turn on cameras or answer questions made it more comfortable to toss thoughts into the chat and speak during discussions."
"The training was very helpful and engaging."
Bring this session to your team
Half-day. Tool-agnostic. Judgment-first.
- What AI is, and is not.
- Where it helps, where it does not.
- How to catch confident, wrong output.
- A decision check every person can apply.
Works with Microsoft Copilot, Claude, ChatGPT, or whatever you already use.
Book this sessionCommon questions
FAQ.
Do participants need any AI experience?
No. The session is built for people who are not power users, including those who have avoided the tools entirely. Existing power users still get value from the judgment and verification frameworks.
Which AI tool do you use?
It is tool-agnostic. We work with whatever your organization already has, whether that is Microsoft Copilot, Claude, ChatGPT, or a combination. The skills transfer across all of them.
How long is it, and what is the format?
A half-day working session, roughly three hours. It is highly interactive: group exercises, live practice on real tasks, and open debriefs. It can be delivered in person or virtually.
Is this technical training?
No. It is about practical use and judgment, not the underlying technology. Nobody leaves needing to understand how a model works. They leave knowing how to use one well and when to question it.
Do you need access to our data or systems?
No. Exercises use curated and sample scenarios. Nothing sensitive or proprietary is required, which keeps the session easy to approve and run.
How many people can attend?
The session works well for a single team or a cross-functional group. We will recommend a group size based on your goals when we scope it with you.
What do people leave with?
A personal action plan with a specific task to apply within two weeks, a simple five-question check to run before acting on any AI output, and a starter prompt library tailored to their work.
The tools are paid for. Capability is the gap.
If your team has AI tools but adoption is uneven, this session closes the gap for the people who need it most.