L&D actively uses tools and stays out of decision-making
L&D teams are more engaged with AI than ever before. The team is creating content faster, building courses in hours instead of weeks, and experimenting with chatbots, quiz generators, and translation tools. By most activity indicators, things are moving.
So why are so many L&D leaders still fighting for a seat at the AI strategy table?
Because activity and influence are two different things, and this gap is eroding credibility in L&D.
We surveyed more than 1,700 learning professionals to understand the current state of AI in L&D. 78% of L&D teams said they are not present when budgets and priorities are being determined, instead executing someone else’s vision.
Like abilities, it affects compounds in tissues. Teams currently developing an AI strategy will be evaluated on their results later, while teams left out of the conversation will not be evaluated. What matters is the ability to play a key role at the most important moments.
Here’s what the remaining data shows and what you can do about it before your next stakeholder conversation.
Numbers that should concern every L&D leader
25% of L&D teams say personalization at scale is the primary reason for adopting AI. Less than 4% of companies prioritize performance.
Now think about your next conversation with an executive. If your CFO walked into a room and asked what your L&D AI investments are delivering, what answer would you get? “We’re personalizing the learner experience at scale,” or “We’ve reduced new hire productivity by 30%. Here’s the data”?
Personalization without a business case doesn’t work well in executive conversations. L&D tends to speak in the language of learner experience, while executives tend to talk about revenue, retention, and productivity, and these two languages don’t currently meet in the middle. The price of that disconnect is authenticity.
Try this before your next stakeholder conversation. Take any AI initiative you’re currently running and ask yourself, “What business metrics should this drive?” Learning metrics alone won’t cut it here, so consider things like time to productivity, sales win rate, compliance incident rate, and customer churn. If you can’t name it, that’s the first problem you should solve, and the one you can solve before entering the room.
Reframe your initiative around that number and lead accordingly. Instead of “We’re improving the learner experience,” think, “We’re using personalized learning to close the skills gaps that are slowing down the sales cycle.”
It’s the same effort, but a completely different conversation.
The problem with resistance is not what you think, and it doesn’t come from where you think
37% of L&D teams say stakeholder resistance is the biggest challenge to AI adoption. Only 12% said a lack of in-house expertise was a barrier.
The resistance that most L&D leaders deal with rarely comes from a single direction. It often occurs from multiple sources at the same time for different reasons. Treating this as one problem with one solution is why so many teams hit the same wall.
Think about who is actually opposing you in your organization right now.
Senior leaders who have never seen the business case they believe in are not anti-AI. They are weighing the risks, but no one has yet shown evidence that the returns justify the investment. It’s a credibility issue, but evidence will solve it.
Managers who don’t believe that AI-generated content meets their team’s standards have probably seen something off the mark or heard enough about AI hallucinations to be careful. This is a quality concern that is addressed by demonstrating a review process.
Employees who are anxious about what AI means for their jobs are more willing to learn. They resist introducing a version of AI that feels imposed on them rather than one designed for them. This is a change management problem that is solved by involving them early, being transparent about what the AI will and won’t change, and making the learning experience feel like development.
Subject matter experts who feel ignored when AI drafts content they previously owned are not saboteurs. They are protecting something important. This is a shared ownership issue that is solved by repositioning them as expert reviewers and quality filters rather than sidelining them.
IT and legal teams, who are slowing things down due to governance concerns, aren’t holding back. They alert you to gaps in the process, and bringing them on board as a partner can help you resolve issues before approval is required.
Importantly, not all of these are equally common within an organization. Diagnosing where the resistance is coming from is really the first step before deciding how to respond. Teams that treat all concerns the same way often default to more communication, more AI training, and end up getting frustrated because they’re applying the right answer to the wrong question.
This is a tactic that works regardless of where the resistance lives. Go to the most resistant person in the room, and it can be whoever it is for you. Ask yourself one question: “What does success look like to you?”
Skip the “What are your concerns” that invites a list of counter-arguments. Also, skip “Show me what the AI can do” which causes a defensive attitude. Leave the question as is and create the next pilot that does exactly that. When skeptics help define success criteria, they shift responsibility from judges to co-owners.
This dynamic works whether the skeptic is a CFO, a line manager, a nervous employee, or a subject matter expert worried about their role.
Resistance usually comes back to trust, evidence, and a sense of control. When you give it to people in a form that is most relevant to their particular interests, resistance tends to move.
The market is fragmented, and the gap is already wider than expected.
27% of L&D teams have been using AI for years, 46% have recently started, and 27% have never started.
Reading this as a gradual curve of early adopters, mainstream, and laggards misses what’s really going on. This is a chasm, and the distance between the groups is increasing every quarter.
The teams with the longest track records already have the lead and continue to extend that lead. Every pilot builds organizational knowledge, every win earns more budget and more permits, and with each quarter of execution the difference becomes harder to close.
A further detail is where the team is investing its AI efforts. The most common uses are content creation (30%) and research (21%). The least common are enhanced reporting (11%) and streamlined delivery (11%). While teams are focusing their AI efforts on the parts of the work that feel familiar, like drafting content and summarizing research, there is a lack of investment in the parts that actually change the strategic position, like connecting learning to outcomes, delivering it when and where it’s needed, and proving its effectiveness.
Using AI to do the same thing faster increases efficiency. Using AI to tackle fundamentally different problems is a strategic shift: one buys time, the other gains influence.
If you’re among the 46% who recently started, take the next step. Choose one of the hottest business problems in your organization right now, such as a new product launch, customer retention crisis, or compliance deadline, and build one AI-assisted learning intervention around it. Measure against business metrics from day one. Concentrated wins in high-profile areas will help you improve your strategic position more than 10 efficiency improvements running quietly in the background. Start small, but start where people will pay attention.
The exclusion cycle and how to break it
Only 22% of L&D teams are participating in AI strategy discussions.
AI is reshaping how organizations hire, develop, and retain talent, but in 78% of organizations, the functions responsible for building capability are left out of the conversation.
This cycle runs as follows: L&D is not included in strategy discussions and cannot shape the direction of AI adoption. If you don’t sit at that table, you can’t conduct experiments that produce evidence. Without evidence, you can’t insist on including it. The cycle continues.
Breaking it means providing proof before the invitation arrives. Proof requires access, and access requires a wedge, so find your wedge.
Find a business leader in your organization who is currently losing sleep over talent issues. It could be a skills gap that impacts delivery times, a new system that no one knows how to use, or a team that continues to meet its goals. Rather than learning a solution, approach them with the question, “Can we run a six-week pilot to help with this? Can we agree upfront on how we’ll know if it works?” Most people would say yes. After 6 weeks, the data were available. Data is a way to understand the content of the conversation. Make exclusions appear like business risks one outcome at a time.
The ethical gap that no one talks about
15% of learning professionals feel prepared to manage the ethical implications of AI in learning.
AI is already informing learning and talent strategies, influencing who gets development opportunities, which learning pathways are recommended, and how performance is assessed. However, the majority of L&D professionals feel unequipped to manage the risks involved.
Organizations that don’t think carefully about bias in AI-generated content, transparency in algorithmic decision-making, and data privacy in learner analytics will not avoid ethical risks. They are postponing it. Ethical risks deferred persist. It lurks quietly until something comes to the fore publicly that is very difficult to reverse.
You don’t need a complete ethical framework from day one. There are three things you need. First is the review step in all AI content workflows. In this step, a human checks the content each time before it is provided to the learner. The second is a clear internal answer to the question, “What learner data are we using and who has access to it?” Third, talk to your legal or compliance team before you scale up, rather than after an issue occurs. These three do not cover every ethical scenario created by AI, but they provide a solid foundation on which to build.
What the data really says
Even if you strip away all the statistics in this article, the story remains consistent. L&D is capable, but it’s not always located where the business needs it.
This gap boils down to the distance between optimizing the learner experience and driving business outcomes. It also shows up in how we use AI, whether it’s to move faster through familiar tasks or tackle more strategic problems.
Teams closing the gap are running one small experiment, measuring the right things, building trust one thing at a time, and building each win into the next.
All actions in this work are single. One metric, one question, one pilot, one wedge, one review step. That’s intentional. Teams that try to solve the AI shift all at once tend to suffer from analysis paralysis, whereas teams that pick one thing and prove it works build a significant compounding advantage.
All AI strategies in L&D can wait. What you need to do now is to intentionally take the next action.
Share with
