AI Champions Program: How to Choose Them and Give Them the Time

An AI champions program names one person in each team to help colleagues use AI well. Champions answer questions, share what works and raise problems with leaders. Choose them for curiosity, patience and the respect of their peers, not seniority or technical skill. Then give them protected time, a clear brief and a direct line to whoever owns AI.

What does an AI champion do?

An AI champion helps the people in their own team get value from AI. They answer everyday questions, show colleagues methods that work, collect good examples, remind people of the data rules and report problems to whoever owns AI. They are a peer guide, not a help desk, a policy owner or an approver.

The role works because it is local. A salesperson asks another salesperson how they draft follow-ups, not a central AI lead they rarely see. Champions translate the company's rules and methods into the language of their team's actual work.

Keep the brief short:

1. Help. Answer colleagues' questions and sit with anyone who is stuck.
2. Share. Show the team one useful method each fortnight.
3. Collect. Gather what works, so it can be written into shared files.
4. Report. Tell the AI owner what is blocking people, every month.

Champions should not approve AI output, grant access to systems or set policy. Those stay with named owners, so the champion is never the person blamed when something goes wrong.

Who makes a good AI champion?

Good champions are curious about AI, patient with colleagues, respected by their peers and careful with risk. They do not need to be the most technical person or the most senior. Often the best choice is someone already quietly helping others, whose team trusts their judgement on everyday work.

Ask for volunteers, then choose. Volunteering signals interest, but a manager's view of who colleagues already go to for help is just as useful. Avoid picking the loudest enthusiast if their team finds them hard to approach.

How much time should AI champions get?

Give champions a fixed block of time each week, agreed with their manager and written into their role, for at least the first few months. Without protected time, the role becomes unpaid extra work. It is squeezed by deadlines and quietly disappears, usually just when colleagues start bringing harder questions.

The time pays back because learning takes hours. According to BCG's 2025 AI at Work survey, regular use is far higher among employees who have had at least five hours of training. A champion can supply part of those hours informally, at the moment someone needs help, which is when it sticks.

Review the time each quarter. Early on, champions answer many basic questions. Later, the work shifts to collecting methods and helping with harder cases. Reduce the time once questions slow, but do not remove it while the team is still changing how it works.

Recognise the role, too. Mention champions' contributions in leadership updates and performance reviews, so the work counts.

How should champions work with leaders?

Champions work best with a direct line to the person who owns AI: a short monthly meeting where they report what is working, what is blocked and what colleagues are worried about. Leaders set direction and remove obstacles; champions bring back the ground truth leaders would otherwise never hear.

Leadership signals still matter most. According to Microsoft and LinkedIn's 2024 Work Trend Index, power users are 61% more likely to have heard from their CEO about using AI at work. Champions extend that signal into each team, but they cannot replace it.

Support is thin where it is needed most. According to BCG, only about a quarter of frontline employees say they get enough leadership support on AI. A champion network is one practical way to close that gap, because it puts help inside the team rather than at head office. The AI adoption playbook shows where champions sit in the wider plan.

How do you set up an AI champions program?

Set it up in six steps: name the executive who owns AI, choose one champion per team, write a one-page brief, protect their time, start a monthly meeting with the AI owner and give them a shared place to record what works. Then launch champions alongside team training, not before it.

1. Owner. One executive accountable for AI, whom champions report to.
2. Selection. One champion per team, chosen from volunteers and peer reputation.
3. Brief. One page: what they do, what they do not, and the data rules.
4. Time. A weekly block, agreed with their manager.
5. Rhythm. A monthly champions meeting, plus a monthly clinic open to everyone.
6. Record. Methods that work are written into shared files, not kept in chats.

Champions should take part in the first team workshop, so colleagues see them as the person to ask next. The guide to running an AI workshop covers that session.

How do you know an AI champions program is working?

It is working when weekly use of the shared AI setup rises in each team, questions move from basic to advanced, and methods champions collect end up in shared files that others use. It is failing when champions stop reporting, colleagues still ask the same basics, or the role keeps changing hands.

Watch for a champion carrying the whole team. If everyone sends them work to run through AI, the team has not adopted anything; it has outsourced to one person. So the fix is more hands-on practice, not more champion time.

The monthly clinic is where problems surface early. The guide to keeping AI momentum covers how to run one.

Skein offers adoption as one of three ways in, alongside strategy and implementation, and measures adoption as one of its outcomes; the AI facilitation page covers how teams learn on their own work. That suits a business with a few keen AI users and many who have not started. A team of five may not need formal champions at all.

Frequently asked questions

What is an AI champion?

An AI champion is an employee who helps colleagues in their own team use AI well. They answer questions, share methods that work, remind people of the data rules and report problems to the executive who owns AI. It is a peer role, done alongside their normal job with protected time.

How do you choose AI champions?

Choose people who are curious about AI, patient with colleagues, respected by their peers and careful with risk. Ask for volunteers, then check with managers who colleagues already go to for help. Seniority and technical skill matter less than trust. Avoid anyone already overloaded or hard to approach.

How much time should an AI champion spend on the role?

Agree a fixed weekly block with their manager and write it into their role, at least for the first few months. Without protected time the role becomes extra work and fades. Review the time each quarter, reducing it once basic questions slow down but keeping some while habits are forming.

How many AI champions does a business need?

One per team is a good starting point, so every group has someone nearby who understands its work. Very small teams can share one. Larger teams with very different roles, such as sales and customer success, usually need separate champions because their AI problems differ.

Should AI champions be paid extra?

Extra pay is not essential, but recognition is. Protect their time, count the work in performance reviews and mention their contributions in leadership updates. What fails is treating the role as unpaid extra effort on top of a full workload, because people quietly drop it when deadlines arrive.

Related reading