The short answer
AI enablement is the ongoing work that makes it possible for employees to use AI well in their own jobs. It covers approved tools, clear rules, training on real tasks, a network of AI champions and measurement of what changes. It’s different from sales enablement, which prepares sales teams to sell and sometimes uses AI to do it. AI enablement is for the whole organization. My advice is to run it as a standing function with a named owner instead of a one-off launch, and to give your champions real time in their calendars.
See how to set up an AI champions program- What Tools, rules, skills, champions and proof, run as one program.
- Who An owner, an executive sponsor, managers and one champion per team.
- First step Name an owner and a sponsor, take a baseline and agree one page of rules.
I advise organizations on AI adoption and run AI change programs, so I have a stake in this topic. I don’t sell software licenses. Sources checked 1 November 2026.
What is AI enablement?
AI enablement is what an organization does so that people can use AI in their daily work, and keep using it as the tools change. I think of it as a function, a bit like onboarding. It never finishes, because new people join, new features arrive and the rules need updating.
The term is used in two other ways. Much of what comes up in search is about AI for sellers, like IBM’s page on AI in sales enablement, and some IT vendors use it for data and integrations. Here it means the people side. Can people across the organization use AI safely in their work, and does it change how the work gets done?
Adoption is the outcome, people using AI because it helps them. Enablement is the program that gets you there. My AI adoption guide covers the bigger picture and change management, and this page covers the enablement work week by week.
Why aren’t the tools enough on their own?
A license gives people access. It doesn’t tell them what to use the tool for, what they may paste into it or who to ask when they get stuck. In BCG’s 2026 AI at Work survey (11,749 workers in 14 markets), 80 percent of the regular AI users with a clear AI strategy but limited access to tools reported measurable business impact, against 60 percent of those with strong tools but limited strategic clarity. These are the respondents’ own reports (BCG).
Managers matter as well. In Gallup’s tracking, employees in AI-adopting organizations who strongly agree that their manager actively supports their team’s use of AI are 1.7 times as likely to use it a few times a week or more (Gallup). Enablement covers what the license leaves out, and the five parts below are how I’d organize it.
What does an AI enablement program contain?
I’d build it from five parts. Tools, rules, skills, champions and proof. It’s my own proposal, so adjust it to how your organization works.
| Part | Minimum version | Owner | Sign that it works |
|---|---|---|---|
| 1. Tools | One approved tool per main need, work accounts and a simple way to request a new tool. | IT, with the enablement lead | People stop asking which tool they’re allowed to use. |
| 2. Rules | One page on which data never goes in, who checks output and which tools are approved. | Legal, IT security and HR | Questions like “can I paste this?” get an answer within two working days. |
| 3. Skills | Short sessions on each team’s own tasks, leaders and managers first. | HR or L&D, with the champions | Each team has two or three documented ways it uses AI. |
| 4. Champions | One named person per team, with agreed time and deeper training. | The enablement lead | Colleagues bring their questions to the champion before they try a workaround. |
| 5. Proof | A baseline, a few measures and a quarterly review that decides what to change. | The executive sponsor | Leaders change priorities based on what they learn. |
I’d rather have all five in a light version than one of them perfect. A champion network without approved tools turns into a support group for shadow AI, and without proof the program is likely to lose its budget at the next planning round. For choosing the tool, see Copilot vs ChatGPT. For the order of training, see the AI literacy playbook.
In the EU, the skills part is also a legal duty. Under Article 4 of the AI Act, providers and deployers must take measures to support AI literacy among their staff and others who use AI on their behalf. According to the European Commission, the duty remains after the Digital Omnibus, in force since 27 July 2026, and no specific level is mandated (European Commission).
In the AI & HR Barometern 2026, humAIn Academy’s survey of 329 HR professionals in Sweden, 61 percent have had formal AI training from their employer, but only 25 percent training that covers law, risk or the AI Act. Of the 201 with training, 75 percent feel confident using AI safely and effectively in HR, against 57 percent of the 127 without. That’s a correlation, and it doesn’t prove the training caused it. My advice is to put the one-page rules into the first session, so people learn the tool and its limits together.
Source: AI & HR Barometern 2026, humAIn Academy (overview in State of AI in HR). 329 HR professionals in Sweden who chose to respond, 8 June to 24 September 2026, probably more interested in AI than HR overall. The figures describe HR respondents, not employees in general. Confident means 4 or 5 on a five-point scale. The 25 percent, the 19 percent below and the training comparison are my own calculations from the survey data.
Who owns AI enablement?
Someone has to own it by name. In the AI & HR Barometern 2026 (329 HR professionals in Sweden, see the note above), 33 percent of the respondents say the leadership team has the main responsibility for AI questions, 26 percent say IT, 23 percent say it’s shared and 6 percent that it’s unclear. Only 4 percent say HR, and 19 percent name unclear ownership among the biggest obstacles. My view is that IT should own the tools, HR or L&D should run the enablement and a business leader should sponsor it. When enablement sits with IT alone, it tends to become a rollout. When it sits with HR alone, it tends to become a course catalog. The bigger questions, such as who approves new tools and which uses count as high risk, belong in an AI governance framework.
| Role | What they do | Time (my suggestion) |
|---|---|---|
| Executive sponsor | Sets direction, uses AI in the open, chairs the quarterly review. | 1 to 2 hours a month |
| Enablement lead | Owns the plan, runs the champion network and reports progress. | Part-time to full-time, by size |
| Champions | Help their team find use cases and pass on blockers. | About half a day a week for the first 90 days |
| Managers | Raise AI in team meetings and decide where saved time goes. | A few minutes in each team meeting |
| IT, security and legal | Approve tools, keep the rules current, answer questions quickly. | A named contact in each |
Managers decide what happens with the time AI saves. In BCG’s survey, 66 percent of non-manager office employees who use AI regularly (BCG’s frontline employees) got limited or no guidance on what to do with that time (BCG). More in my guide for leaders.
AI enablement lead role description (template)
| Field | What to write |
|---|---|
| Purpose | Help people across [organization] use AI safely in their daily work, and keep it working as the tools change. |
| Reports to | [Executive sponsor], with named contacts in IT, legal and HR. |
| Responsibilities | Run the 90-day plan, the champion network and the quarterly review. Keep the rules and the list of approved tools current with IT and legal. Plan training by team. Share use-case cards across teams. |
| Not part of the role | Buying or configuring tools (IT), the legal basis for data use (legal) and doing each team’s AI work for them. |
| Time | [Share of a full-time role], agreed with [executive sponsor] and reviewed after 90 days. |
| After 90 days | A champion and two or three use-case cards in each first-wave team, a baseline with a first comparison, and a decision on the second wave. |
How do you set up an AI champions program?
An AI champion is an employee who helps the rest of their team use AI in their everyday work. Champions matter most when people know how to use the tool but drop it when the day gets busy, which is the Ability step in ADKAR. These are the steps I’d follow.
- Decide what champions do and don’t do. Champions find and share use cases, answer first questions and pass on blockers. They don’t approve tools, write policy or fix integrations. Write that down, or the role grows until nobody wants it.
- Pick the people colleagues already trust. Look for the people others ask for help, and who are open about what didn’t work. Aim for one per team that shares the same work, since a champion in finance can’t help recruiters. OpenAI suggests starting a network with 3 to 5 credible people (OpenAI Academy). I’d start with one per team instead, in 5 to 15 teams, which is still small enough for one lead to run a weekly huddle with everyone.
- Agree the time in writing with their manager. I suggest about half a day a week for the first 90 days, written into their objectives. Otherwise it becomes unpaid overtime, which I see as the biggest risk to a champion network. Put the cost in the business case. Ten champions at half a day a week is about one full-time person for 90 days. Review the time with each manager after that.
- Train the champions first, and deeper. Give them earlier access, extra sessions on their team’s tasks and a walk-through of the rules.
- Set a rhythm. Hold a 30-minute huddle every week for the first 90 days, then every other week. Add one shared channel and a monthly show-and-tell in each team’s normal meeting.
- Give them a fast lane. They need named contacts in IT, legal and HR who answer within two working days, and a reply on every blocker that says who owns it.
- Recognize the work and rotate. Name champions in leadership updates, include the role in performance reviews and let people step down after a year. Microsoft’s adoption guidance suggests a one-page champion charter that sets out the expected time commitment and recognition (Microsoft Adoption).
AI champion role description (template)
| Field | What to write |
|---|---|
| Purpose | Help colleagues in [team] use AI safely in their daily work. |
| Responsibilities | Find and document two or three use cases with the team. Answer first questions. Run a 15-minute show-and-tell once a month. Pass on blockers to [enablement lead]. |
| Not part of the role | Approving tools, writing policy, technical support and checking colleagues’ output. |
| Time | About half a day a week for 90 days, agreed with [manager] and reviewed on [date]. |
| Recognition and term | Part of the champion’s objectives for the period, and named in leadership updates. One year, then the champion decides whether to continue. |
The 30-minute weekly huddle (template)
- 5 minutes. What’s new in the tools or the rules this week.
- 10 minutes. One champion shows one use case from their team, including what went wrong.
- 10 minutes. Blockers. What stops people, and who takes each one.
- 5 minutes. What each champion takes back to their team.
A use-case card (template)
Give every use case the same card, so other teams can copy it. For the prompt field, my guide on how to write a good prompt has a template the champions can start from.
| Field | What to write |
|---|---|
| Task | The task in the team’s own words, such as “turn meeting notes into a project update”. |
| Tool and prompt | The approved tool, and the prompt or steps ready to copy. |
| Data rules | What may and may not go in for this task. |
| Who checks | Who reviews the output before it’s used. |
| Time before and after | A rough estimate, tested on a few real cases. |
A 90-day plan for AI enablement
This plan sets up the function and the first wave of champions, assuming you have one approved tool. To test a single use case first, see how to run an AI pilot.
- Weeks 1 to 2. Owner, baseline and rules. Name the enablement lead and the sponsor. Take a baseline from usage reports and the short anonymous survey below. Agree the one-page rules.
- Weeks 3 to 4. Champions and workflows. Recruit the first champions together with their managers and agree the time in writing. Pick two or three workflows per team. Before anyone is trained, time three real cases of each workflow and note the quality. That’s your before figure. Train the champions.
- Weeks 5 to 8. Teams on their real tasks. Train each team on its own workflows, leaders and managers first. Start the weekly huddle and the shared channel. Write the first use-case cards.
- Weeks 9 to 12. Measure and decide. Repeat the survey and the timed cases, and compare with the baseline. Hold the first quarterly review. Decide what to scale, what to drop and who joins the second wave. Thank the champions in public.
Train by team, so the people who share the work learn together. When colleagues at humAIn Academy and I worked with Mediemyndigheten, the Swedish Agency for the Media, we ran three two-hour sessions for five groups, about 80 employees, grouped by department. Managers had two sessions of their own.
Baseline survey in five questions (template)
- How often do you use an approved AI tool in your work? Daily, weekly, monthly, rarely or never.
- Which tasks do you use it for? Free text.
- What stops you from using it more? I don’t know what to use it for, I’m unsure what’s allowed, I have no time to learn, the tool doesn’t fit my work, or something else.
- How confident do you feel using AI safely and effectively in your work? 1 to 5.
- Which task would you most like help with? Free text. The answers become your champions’ first use cases.
Run it anonymously in weeks 1 to 2, in weeks 9 to 12 and then every quarter. Count 4 or 5 on question 4 as confident. Confidence isn’t the same as skill. To see what people do on a real task, add the 30-minute work sample from the AI fluency framework.
How do you measure AI enablement?
In the AI & HR Barometern 2026, 70 percent of the HR respondents say their organization doesn’t measure the effect of AI in HR, and 5 percent use defined KPIs. I start from the three signals in my AI adoption guide (breadth, depth and momentum), make them countable and add three that belong to enablement.
| Metric | What to measure | Where the data comes from |
|---|---|---|
| Breadth | Teams that use AI in real work, and the share of people who use an approved tool at least weekly | The tool’s usage reports and the use-case cards |
| Depth | Documented use cases on meaningful tasks, per team | The use-case cards |
| Momentum | New use cases that teams suggest themselves, per month | Notes from the champions’ huddle |
| Effect | Time and quality on two or three named tasks, against the baseline | Before-and-after tests on real cases, timed in weeks 3 to 4 and again in weeks 9 to 12 |
| Confidence | Share who answer 4 or 5 on the confidence question | The baseline survey, every quarter |
| Champion health | Champions still active, and hours spent against hours agreed | The champions and their managers |
Agree before you start which results would make you change course, and who decides. Here are three rules I’d write down before the first review.
- If fewer than half of the champions are still active at week 12, fix the time agreements before you recruit a second wave.
- If a team has no use-case card after 90 days, its champion and manager pick one task together and test it in the next month.
- If a use case saves no time and doesn’t improve quality, drop it and tell the other teams why.
Why do AI enablement programs fail?
Start from what you can see, then work back to the cause.
| What you see | Likely cause | What to do |
|---|---|---|
| Huddle attendance drops after a few weeks | Champions have no agreed time | Go back to their managers, put the hours in writing and check them in the quarterly review. |
| Champions answer tool questions, but few use cases appear | Champions were picked for technical skill | Ask each team who they go to for help, and add that person. |
| People attend training, but their work doesn’t change | The training had no task | Rebuild the sessions around the team’s own workflows. |
| Teams use AI, but managers never mention it | Managers are on the sidelines | Give managers their own session and a standing AI question in team meetings. |
| The program feels stale after six months | It was run as a launch with an end date | Keep a named owner and the quarterly review. |
Frequently asked questions
What is the difference between AI enablement and sales enablement?
Sales enablement prepares sales teams to sell, and AI in sales enablement means using AI for that work. AI enablement helps everyone in the organization use AI in their own jobs, from HR and finance to operations.
What is an AI champions program?
An AI champions program is a network of employees who help the rest of their team use AI in their daily work. Champions find and share use cases, answer first questions and pass on what blocks people. They need agreed time and deeper training.
What does an AI enablement lead do?
An AI enablement lead owns the plan for helping people use AI in their work. They run the champion network, keep the rules and approved tools current with IT and legal, plan training by team and report to the executive sponsor every quarter. In a small organization it can be part of an HR or L&D role. In a large one I’d make it full-time.
How many AI champions do we need?
Count teams. You want one champion for each group that shares the same work, so 20 such teams need about 20 champions over time. Start with a first wave in 5 to 15 of the teams most ready to go, and add the rest later.
Can a manager be an AI champion?
It can work, but I’d pick peers. People tend to ask a colleague things they’d hesitate to ask their manager, such as what they’re allowed to paste. Managers have their own job in enablement, raising AI in team meetings and deciding where saved time goes.
Sources
- BCG: AI at Work 2026, Strategy Matters More Than Tools (slideshow, June 2026)
- BCG: AI Is Reshaping Jobs Faster Than Companies Are Reshaping Work (press release, 3 June 2026)
- Gallup: Global Indicator, Artificial Intelligence (data as of May 2026)
- State of AI in HR (AI & HR Barometern 2026, humAIn Academy)
- OpenAI Academy: Build and grow a network of local AI Activators
- Microsoft Adoption: Define your AI Champion role and charter
- IBM: AI in sales enablement
- European Commission: AI literacy questions and answers (updated 20 October 2026)
- European Commission: Navigating the AI Act (Digital Omnibus in force 27 July 2026)
Want help setting up AI enablement?
In an AI change program, I work with your leaders, managers and HR for 3 to 12 months on training, responsibilities and follow-up. Many organizations start with a 3-hour leadership workshop from SEK 24,900 excluding VAT. The leadership teams I’ve worked with give an average rating of 4.97 out of 5, measured across 13 groups.