AI’s Greatest Impact on Learning isn’t Content but Decisions

University of Cambridge’s Paul Clothier offers this commentary on how AI’s greatest impact on learning isn’t content but decisions. This article originally appeared in Insight Jam, an enterprise IT community that enables human conversation on AI.

A lot of the excitement surrounding AI in Learning and Development has focused on the ability to create courses, assessments, videos and other learning materials in a fraction of the time they once required. This is understandable. Learning teams are under constant pressure to deliver more, to respond more quickly to business needs, and to do this with limited resources. AI’s ability to accelerate content development is already delivering real value here.

But there’s a risk.

By focusing on AI’s ability to create learning content, we may be overlooking what could become AI’s greatest contribution to Learning and Development.

The question isn’t just, “How can AI help us create better learning content?” It’s, “How can AI help people make better decisions while they work?”

It’s an important shift in emphasis, changing how we should think about the role of AI in workplace learning and, just as importantly, how we should evaluate its success.

The goal of Learning and Development in organizations has never been to produce courses or increase knowledge for its own sake. The real goal has always been to improve workplace performance.

Knowledge matters because it helps people know what to do. Sometimes that means applying a skill. Sometimes it means choosing the best course of action. In both cases, what matters is what people do with what they know.

Content has never really been the destination. It’s just been one of the vehicles for getting there.

This is where AI changes the conversation.

Until now, most learning technologies have focused on helping people access information. Sometimes that information was presented via a classroom, sometimes through e-learning, and increasingly through mobile devices and performance support tools. Each step brought help and support closer to where the work actually happened.

AI has the potential to shift workplace learning from helping to deliver information to supporting better decisions.

Instead of just presenting information, AI can help employees think through a situation before they act. It can ask questions, challenge assumptions, and explore options. And rather than giving everyone the same guidance, it can adapt its support to the employee, their experience, and the decision they’re facing.

Mobile devices brought information to the point of need. AI brings decision support to the point of action.

But this only works if employees treat AI as a thinking partner, not an authority. AI can be wrong. The more important the decision, the greater the risk of simply trusting its answer. Learning and Development has an important role to play in helping employees question AI’s guidance, use their own judgement, and recognise when human expertise is needed.

That doesn’t necessarily reduce the importance of classes, courses, or performance support. Quite the opposite. AI builds on everything we’ve learned over the past two decades. Courses remain an effective way to build foundational knowledge. Performance support remains an essential way of delivering information at the point of need. AI extends these approaches by making support more responsive, more interactive and more relevant.

Think about the kinds of challenges employees face every day. A manager preparing for a difficult performance conversation. A salesperson deciding how to respond to a customer’s concerns. An engineer diagnosing an unfamiliar fault. Or me last week, trying to decide how to phrase a difficult email to a colleague without sounding too harsh.

Information is only part of the solution in such cases. Success depends on choosing the best course of action for that particular situation.

That’s where AI can have its greatest impact.

For Learning and Development teams, this means starting with a different question. Instead of asking where AI can help create learning content faster, we should identify the decisions that have the greatest impact on performance. Where do employees hesitate, make mistakes, or struggle to choose the best course of action? These are the moments where AI-supported guidance may have the greatest value.

The discussion about AI in Learning and Development needs to move beyond productivity alone. Faster course development is valuable. So is reducing the cost of producing learning materials. These benefits shouldn’t be dismissed.

But neither should they become the primary measure of success.

If AI helps employees to respond more effectively to new situations, apply their knowledge more confidently, and choose better courses of action, its contribution will extend far beyond just content creation.

This also has implications for evaluation. Learning organizations have traditionally measured completion rates, assessment scores and, more recently, the efficiency gains AI brings to content development. These measures are useful, but they tell us little about whether people are making better decisions as a result. Evaluation frameworks such as LTEM remind us that performance matters more than knowledge alone, and Tier 5 specifically addresses decision-making competence, exactly the outcome AI-supported guidance is meant to improve.

As learning leaders, we should continue to ask how AI can help us produce high-quality learning content more efficiently. It’s an important question, but it’s not the most important one.

The more important question is this:

How can AI help employees make better decisions at the moments that matter most?

That, I believe, is the question Learning and Development should now be asking.