Who should own AI on your team?

Hi Everyone,

Someone on your team is already massively ahead of everyone else on the AI front. You just haven't recognized it yet.

A big part of the reason is decentralized, or independent, usage. Microsoft's 2024 Work Trend Index found that 78% of employees using AI at work are bringing their own tools.

Today, we're covering how to find the right person to lead AI adoption on your team, how many you need, and why the answer probably isn't someone in IT.

Why the IT instinct usually misses the mark

When leadership asks, "Who should own AI for us?", the default instinct is often to hand it to the most technical person available. But the strongest AI champions are rarely in IT. More often, they sit in finance, operations, or marketing — because they understand their team's workflows well enough to spot where AI can save time and remove friction.

These champions succeed because peers trust them and because they know how the work actually gets done. Whether they can explain how a model works is largely irrelevant. What matters is that they can show a colleague a better way to draft a client memo, prepare for a meeting, or automate a repetitive task using AI.

This shift is also changing leadership expectations. A recent survey of senior leaders found that 81% have raised their expectations of midlevel managers when it comes to adopting and integrating digital tools over the past year.

How many champions do you need inside your business?

Published ratios vary widely, from one champion per 15 employees to one per 50. A common starting point is roughly one per 25, spending three to four hours a week.

Citi is the clearest proof the model works at scale. The bank built a network of about 4,000 "AI Accelerators" (colleagues from across operations, risk, customer support, and technology, not a single specialist team) and firm-approved AI tool use is now above 70%. For a 300-person company using the one-per-25 starting point, that's about 12 AI champions, each spending half a day a week supporting best practice usage and adoption.

Training pays. In DataCamp's 2026 survey, companies that paired AI tools with structured training reported significant AI ROI at roughly double the rate of those that didn't (42% vs. 21%).

In those 3–4 hours a week, champions usually focus on a handful of tasks:

• Run a 30-day pilot on one workflow and track the time it saves
• Share prompts and give quick demos in team meetings
• Help colleagues one on one when they get stuck
• Show the results each quarter to a senior sponsor

Spotting your first champion

Peer-to-peer learning is the top way employees build AI skills, with 69% ranking it in their top three in a recent survey. That means your champion needs to be someone peers listen to and trust, not just the most enthusiastic early adopter. Look for people who fit these habits.

• They already experiment with AI on their own work and can point to a specific example
• Colleagues already ask them for help with tools or shortcuts
• They focus on fixing process problems, not just trying new tech
• They share what they learn without being prompted
• They know what data can and can't go into AI tools

You're looking for the person who already does this informally. Give them dedicated time and a clear mandate, and the scattered tinkering turns into something you can point to and scale.

Try this today

Pick the function with the most repetitive knowledge work and ask the manager two questions.

• "Who on your team is already using AI to save time?"
• "Who do others go to when they need help with it?"

If the same name comes up twice, that's your first champion.

Give them 3–4 hours this week to document one workflow where AI saves time, then have them show it to the team next Monday. You'll learn more from that single demo than from a month of vendor presentations.

Go deeper

👉 Microsoft Work Trend Index: AI at Work Is Here. Now Comes the Hard Part – The source for the 78% BYOAI stat and the training gap data. Worth reading the full report for the employee behavior findings.

👉 Braincuber: How to Build Internal AI Champions in Your Organization – Detailed playbook covering champion ratios, the Citi case study, and weekly activity frameworks.

👉 BCG: The AI Adoption Puzzle — Why Usage Is Up But Impact Is Not – Strong data on why peer learning outperforms top-down mandates.

👉 Cambridge Spark: Why Every Business Needs an AI Champion – Covers behavioral traits for champion selection and why business functions outperform IT as champion sources.

Coming up tomorrow

Tomorrow we'll show you how to build an off-site agenda around the decisions the offsite should help you make, rather than the topics each team wants to present.

That's it for today!

P.S. Have you given anyone an AI champion role yet, formal or informal? Tell us what's worked.