Building internal champions and an AI council
How to find the people who'll carry adoption and how to give them enough room.
Top-down AI mandates produce compliance. Peer-driven adoption produces belief. Internal champions are how the second happens.
Who makes a good champion
Not the most senior person. Not the most enthusiastic person (enthusiasm alone breeds cynicism). The right profile:
- Respected by peers (not necessarily liked).
- Technically capable enough to use AI tools credibly.
- Honest about limitations — doesn't overhype.
- Willing to invest time beyond their day job.
- Good at translating — can explain AI value to non-technical colleagues.
Often mid-level individual contributors or senior ICs, not managers.
How to find them
- Observe usage data. Who's using AI tools voluntarily and effectively?
- Ask around. "Who on your team is playing with AI?" usually produces names.
- Open calls with real incentives (time, recognition) attract self-identified candidates.
- Pilot volunteers — people who raised their hand for early trials.
Avoid anointing champions top-down. "You're now the AI champion" without buy-in fails.
The ask
Clearly scope what the role involves:
- Time commitment: typically 2-5 hours/week.
- Activities: lead office hours, write internal docs, train colleagues, report back on friction.
- Duration: 6-12 months (rotational makes sense).
- Recognition and support: explicit; this isn't "extra" work piled on.
Without clarity, champion roles become volunteer burnout.
The AI council
For larger orgs, champions should organize:
- Monthly cross-team meetings.
- Shared Slack channel for ongoing questions.
- Quarterly reporting to leadership.
- Rotating presentation — each champion shares wins and challenges.
This creates a peer network; champions support each other; best practices spread organically.
What champions actually do
Week to week:
- Help colleagues get unstuck. Specific questions, specific answers.
- Share patterns that work. "Here's a prompt that nailed it for me."
- Flag friction. Report issues to the AI program team.
- Demo real usage. Show-and-tell in team meetings.
- Advocate for governance changes. "This approval process is blocking adoption."
They're the translation layer between the program and the users.
What they shouldn't do
- Be the support desk. Sustainability means users learn to self-serve.
- Be unpaid consultants. Their time needs recognition.
- Defend every vendor decision. Independence from the procurement side matters.
- Be the only voice. Multiple champions per area; no single point of failure.
The power of visible example
Nothing beats "a peer I respect uses this and says it works" for adoption. Champions provide this at scale.
Practical patterns:
- Live demos in team meetings. Real work, real tools.
- Internal write-ups of how champions use AI in their daily work.
- Pair sessions where champions sit with colleagues learning.
These are worth more than any top-down training program.
Compensating champions
Options:
- Recognition. Internal awards, visibility to leadership.
- Training budget. Send them to AI conferences.
- Protected time. Explicit 10% of their week for the role.
- Stipends. Real money for real extra work.
At minimum, recognition and protected time. Paying for genuine extra work is appropriate when asks are substantial.
The anti-pattern: AI evangelist
Beware the person who:
- Promotes every new tool.
- Overstates AI capabilities.
- Dismisses concerns.
- Dominates conversation.
They create skepticism, not belief. You want credible enthusiasts, not zealots.
Sustaining the network
Champion programs decay without intentional sustenance:
- Refresh the roster annually — some cycle out, new ones come in.
- Keep leadership engaged — quarterly exec meetings with champions.
- Celebrate success stories publicly.
- Document institutional knowledge — when a champion rotates out, their knowledge stays.
The honest limits
Champion networks aren't sufficient alone. They complement:
- Central program management (strategy, vendor relationships, governance).
- Training infrastructure (formal learning, not just peer).
- Support systems (real helpdesk, not champions alone).
Champions accelerate adoption. They don't replace the operational backbone.
Check your understanding
2-question self-check
Optional. Your answers feed your knowledge score on the track certificate.
Q1.The best AI champions are usually…
Q2.A champion network needs…
Continue in this track
More lessons from The Executive's AI Adoption Playbook.
Lesson 2
Reading your org's AI readiness honestly
A diagnostic that doesn't flatter you — tooling, data, permission, and leadership.
Lesson 3
Picking the right first three use cases
The selection criteria that separate future wins from the ones that stall out.
Lesson 5
Shaping workflows, not just handing out tools
The biggest adoption mistake is buying tools without changing how work actually gets done.