Most companies have done the easy half of AI. The licenses are bought, and usage is climbing. The other half is what decides whether anything changes: people knowing what they may do, learning on their own work, having someone to ask, and someone deciding where the saved time goes. That half has a name, AI enablement, and in most mid-size companies nobody owns it.
What is AI enablement?
AI enablement is the ongoing work of giving people what they need to use AI well in their own jobs: approved tools, rules they can follow, training on their own tasks, someone to ask and a plan for the time AI frees.
This guide is for leaders who own AI at companies of roughly 20 to 1,000 people: COOs, CEOs and founders, HR and People leads, IT leads, and partners at professional firms. It covers what enablement includes, why it matters more than the tools, who should own it, how to measure it and a 90-day plan. If you're planning the wider rollout, our guide to AI deployment covers the stages around it.
One note on the term. Search results for "AI enablement" mix in sales enablement, which is a different job: giving sales teams content and coaching. This article is about the company-wide meaning.
AI enablement vs AI adoption vs AI training
The three get used as if they meant the same thing. The difference decides who owns what.
| AI adoption | AI training | AI enablement | |
|---|---|---|---|
| What it is | Giving people AI tools | Teaching people to use them | Everything that makes use safe and useful |
| Typical output | Licenses, seats | Courses, sessions | Rules, training, support, measures |
| Usual measure | Seats and logins | Completions | Changed work and time put to use |
| Who usually owns it | IT or procurement | HR or L&D | One named lead, with IT, HR and managers |
| Where it stalls | Usage without results | Generic courses nobody applies | When nobody owns it |
Microsoft makes the point about adoption from its own rollout: "a tool licensed to over 200,000 people does not change how the work gets done." Training on its own doesn't close the gap either. In the same post, Microsoft says an early pilot "taught us that AI transformation is a team sport and that tools and training alone were not enough."
Why does AI enablement matter more than the tools?
Because use has outrun support: most staff now use AI, but few get clear direction, enough training or a plan for the time it saves.
BCG's fourth annual AI at Work survey, published on June 3, 2026, covers 11,749 workers across 14 markets. It found that 74% of frontline employees now use AI every day or a few times a week, up from "only around 50%" in earlier years. Among frontline employees who are regular users, 42% report saving eight hours a week. But "66% still receive limited or no guidance on what to do with the time they save." And while 72% of all respondents say the skills expected of them have shifted, "only 36% feel that they have received adequate upskilling."
BCG's conclusion is that direction matters more than access. In its words, "Clear strategy lifts measurable business impact by 25 percentage points," while "better tools, without that strategy and redesign, move it only by approximately 5 points." Only a third of frontline employees say leadership's communications about AI are clear.
Gallup's numbers for the US point the same way. As of May 2026, 30% of US employees use AI at work a few times a week or more, but only 25% say their organization has communicated "a clear plan for integrating AI into current practices." In organizations bringing AI in, 36% strongly agree that their manager supports their team's use of AI. Those who do are 1.7 times as likely to use it a few times a week or more.
The cost of skipping enablement is that use goes where you can't see it. In BCG's 2025 survey, more than half of employees said that when they don't have the AI tools they need, they "will find alternatives and use them anyway." That's shadow AI, and our guide to shadow AI covers what it puts at risk.
What does AI enablement include?
Five parts: approved tools, rules people can follow, training on their own work, someone to ask, and a plan for the time AI frees. Leave one out and the others weaken.
Approved tools and access
People need to know which tools they may use for work, with which accounts, and how to get access. A short list beats a long one. Start from what's already running, including AI features vendors have switched on inside software you already pay for. The free Shadow AI Audit builds a first list in about five minutes, and our lesson on approved tools covers the checks before a tool goes on it.
Rules people can follow
Staff need a few rules they can remember without opening a policy: what data can go into which tool, what AI-assisted work must meet before it goes out, and what to do when something goes wrong. These are the guardrails in our framework. If you don't have a policy yet, the free AI usage policy generator drafts one from your answers.
Training on their own work
Generic AI courses teach prompts. People change their work when they practice on their own tasks, in time set aside for it. BCG's 2025 survey found that "Regular usage is sharply higher for employees that receive at least five hours of training", and that "Only one-third of employees say that they have been properly trained." Start everyone with a shared baseline, such as our free first-hour course, then train each team on its own work, as our lesson on training people on their own work describes. If you want this done for you, that's our employee AI training service.
Someone to ask
Questions come up in the work, not in training: "Can I put this contract in?", "Is this output good enough to send?". Enablement gives them one place to go and a promise of an answer. Our lesson on the front door and the 10-day answer sets one up. Peer support does the rest. Microsoft found that "Weekly peer-led huddles turned experimentation into habit and scaled best practices to everyone on the team." Your most experienced users are the obvious hosts, and what your power users are telling you shows how to find them.
A plan for freed time
This is the part most companies skip, and BCG's 66% figure shows the cost. "The time that individuals save leaks out of the organization unless it is tracked and deliberately reinvested," BCG writes. Decide where freed time goes before it appears: more client work, a backlog, training, or fewer hours. Our lesson on the goal, and where freed time goes walks through the choice.
Who should own AI enablement?
One named lead owns enablement day to day, a leadership sponsor backs it, and IT, HR and managers each hold one part. In a company of 20 to 1,000 people the lead is usually someone in operations with part of their week set aside, not a new hire.
Splitting it this way matches what each function already does well:
- The enablement lead keeps the five parts working together, runs the question channel and reports each quarter on what's changed.
- IT and security keep the approved tools, access and data rules.
- HR and People run training and update roles and job descriptions as the work changes.
- Managers show their teams how they use AI, protect learning time and decide with each team where freed time goes.
- The sponsor on the leadership team sets the aim, signs off the rules and holds the lead to the measures.
Managers matter more than any other group. Microsoft reports that "when managers actively model AI use, reported value from agentic AI rises 17 points and trust in it rises 30 points." In BCG's 2025 survey, the share of employees who felt positive about generative AI rose from 15% to 55% with strong leadership support, yet "Only about one-quarter of frontline employees say they receive that support."
Do you need an AI enablement team?
Not at first. A lead with protected time, a few experienced users acting as the first point of contact in their teams, and the IT and HR owners above are enough for most companies under 1,000 people. A dedicated team makes sense once the question channel, training and tool reviews take more than one person's week. Our lesson on money, time and who's responsible helps set the lead's time and budget.
How to set up AI enablement in 90 days
The steps below give you the five parts, with owners, inside a quarter. Each one produces something the next step needs.
01 · Name the lead and the sponsor
Write down who owns enablement, how much of their week it gets, and which leader sponsors it. Without this, the rest becomes a set of documents nobody keeps current. This step produces the names the rest of the plan hangs on.
02 · List what's already in use
Before you approve anything, find what's running: personal accounts, team subscriptions, and AI features inside existing software. The Shadow AI Audit is a quick start. The list shows you which tools to approve first and which teams are already ahead.
03 · Publish the approved tools and the rules
Turn the list into a short approved list with how to get access, and a one-page set of rules: data, output standard, and what to do when something goes wrong. Make it safe to say how you use AI today, so people move onto approved tools instead of hiding what they use. Our lesson on running a disclosure amnesty covers how.
04 · Give everyone the same first hour
Put every member of staff through the same baseline, so the rules are known and not just published. The free first-hour course ends with a completion record you can keep. This step produces a trained baseline and the first questions for the channel.
05 · Open the question channel and train each team on its own work
Launch one place to ask, with a promised answer time, and a short weekly session where experienced users show their own work. Then pick one team and train it on its real tasks in protected time. That team's results become the example for the next.
06 · Decide where freed time goes and report it
With the first team, agree where the time AI saves will go, and pick one business measure for their work, such as cycle time or rework. At the end of the quarter the lead reports what changed to the leadership team and picks the next team. When a team's results show a whole workflow needs redesigning, that's AI transformation, and the enablement you've built is what it runs on.
How do you measure AI enablement?
Measure whether the work changed, not whether the licenses are used. Seats and logins tell you about adoption. For enablement, track a short set each quarter:
| Measure | What it tells you |
|---|---|
| Share of staff through the baseline training | Whether the rules are known |
| Time to answer a question | Whether the support works |
| Ideas raised, and the share answered | Whether people bring their work to you |
| Teams trained on their own tasks | How far enablement has reached |
| One business measure per team, such as cycle time | Whether the work changed |
| Where freed time went | Whether saved time was put to use |
Keep it to what you'll actually review. A measure nobody looks at is worse than none, because it suggests someone is watching.
Where to get help with AI enablement
Most of the plan above needs your own leaders' time more than outside help. Outside help earns its fee on what you can't staff: training at scale, a policy for a regulated business, or a workflow that needs building. If you want a hand, our AI implementation services cover all three. We sell those services, so weigh that, and know that the plan above works without us.
What's still open
- The evidence comes mostly from large surveys of large companies. BCG and Gallup survey employees across many organizations, and Microsoft's lessons come from one very large company. There's little published data on enablement in companies of 20 to 1,000 people.
- Self-reported numbers. Time saved and training adequacy in these surveys are what employees report, not measured hours.
- Agents change the job. BCG found that half of respondents say their companies lack clear governance for managing teams with people and AI. As agents take on more work, enablement will need rules for them too.
We'll update this article when new evidence on AI enablement, training or manager support is published. The date at the top of the page shows when we last checked.
Sources
- AI at Work: Why Strategy Matters More Than Tools, BCG, June 2026: frontline use at 74%, eight hours saved, the 66% without guidance on saved time, skills expectations and upskilling, leadership communications and governance for teams with AI.
- AI Is Reshaping Jobs Faster Than Companies Are Reshaping Work, BCG, June 2026: the survey's sample of 11,749 workers across 14 markets, and the effect of clear strategy against better tools.
- AI at Work 2025: Momentum Builds, but Gaps Remain, BCG, June 2025: training of five hours or more, the share properly trained, leadership support and employees finding their own tools.
- Artificial Intelligence indicator, Gallup, data as of May 2026: frequency of AI use by US employees, the share whose organization has a clear AI plan, and manager support.
- What we've learned from Microsoft's own AI transformation, Microsoft, September 2026: licenses against changed work, tools and training not being enough, peer-led huddles and managers modeling AI use.