AI in eLearning: It’s Not What You Think Matters Most

There’s a lot of noise about AI in the eLearning world at the moment. Depending on who you speak to, it’s either about to decimate the industry or create a panacea of learning brilliance.

Neither of those takes is especially helpful. What’s been more interesting for us is the reality of using it day to day. Not as a gimmick. Not as a shortcut. But as something that quietly improves how the work gets done.

At Knowledgefront, we’ve ended up embedding AI into pretty much every part of the process. Not just in the courses or the platforms, but in how we actually create learning.

AI Isn’t the Product. It’s the Engine.

When people ask about our AI strategy, they usually mean features.

• Is there a chatbot?

• Is the LMS AI-powered?

• Are the courses using AI?

All of that is interesting - there’s real value in how AI can sit within platforms and learning experiences, and we’re actively exploring it. But as digital learning partners, we’re just as interested in something that gets talked about a lot less.

"How AI enhances the learning design, production, and creative process"

Because that’s where much of the impact lies for our clients. Not just in what the learner sees at the end, but in how effectively, efficiently, and intelligently that learning gets designed and built in the first place, and that’s the bit we spend a lot of time on.

It Starts Earlier Than You Think

Before any scripts or storyboards, there’s always that slightly messy phase where you’re trying to get your head around a topic. Sometimes it’s straightforward. Sometimes it’s a dense policy document, a complex technical subject, and three stakeholders who all describe it differently.

AI has changed that stage quite a bit.

Not because it gives you “the answer”, but because it gives you something to challenge. You can test assumptions, ask for different perspectives, and sense-check what you’ve been told. It’s very good at exposing gaps in your own understanding.

So by the time we get into design conversations, we’re not just reacting. We’re already a bit further along.

AI Gets You to Version v0.6 Very Quickly

The blank page used to slow everything down. Now we’ll generate a rough first pass of a script or structure just to get something moving. It’s not finished, and it’s not supposed to be. It just gives you something to react to.

You can try different tones. Rework it into a scenario. Strip it back. Push it in a different direction. It speeds up the messy middle of the process. But the important bit is what happens next.

We take it to v1.0.

That’s where experience comes in. Knowing what actually works for learners. What sounds natural. What aligns with the client’s voice. What needs simplifying, and what needs more depth. AI gets you started. It doesn’t get you finished.

AI Expands the Sandbox. We Still Design the Experience.

This is especially true with scenarios. It’s very easy to end up with scenarios that feel a bit flat. Either too obvious or not quite believable. AI helps us explore more. Different angles, different outcomes, different “what if” moments. It gives us more raw material to work with.

But it doesn’t understand nuance, context, or audience in the way we do. So we shape it. We turn it into something that feels real, relevant, and worth engaging with.

The sandbox is bigger, but the design is still ours.

Asset Creation: Filling the Gaps Properly

This is where people expect AI to show up, and it does, but not just as a replacement for everything. It’s more about solving the problems that used to slow things down.

A good example is a course we built recently, where we needed a very specific image: a hydrogen bus being refuelled, at a particular angle, showing specific features of the bus. Not a generic “green transport” image. Not a bus near a fuel pump. The actual thing.

Stock libraries didn’t have it, or they had versions that were close, but not quite right. The wrong setup, the wrong context, the wrong visual detail. Previously, we would have had an illustrator create this for us, adding time and cost to the project.

So, using Adobe Firefly we generated it. We created an image that matched exactly what we needed, refined it, and dropped it straight into the course. No compromises, no “that’ll do”. It’s a small detail, but it makes the content clearer and more credible.

The same applies more broadly. We use AI to create visuals that actually match the learning, generate voiceovers, build video where it adds value, and iterate quickly when things change. It gives us more control over the final product, rather than being constrained by what already exists.

Enhancing Courses Without Going Fully Bespoke

One of the more interesting shifts has been on the technical side. We’re not turning every project into a custom-built programme, and most of our work still sits within tools like Rise and Storyline, but we are increasingly using AI to help us generate small snippets of HTML, CSS, and JavaScript that enhance what those tools can do.

Things like:

• more flexible layouts

• lightweight interactive elements

• small visual behaviours or animations

• UI tweaks that improve usability

Previously, you either accepted the tool's limitations or spent significant time working around them. Now we can extend them just enough to improve the experience, without turning it into a full development project.

It’s a useful middle ground. More flexibility, without unnecessary complexity.

A Very Fast Second Pair of Eyes

AI is also surprisingly useful as a reviewer. It’s good at spotting inconsistencies, repeated phrasing, or areas where something isn’t quite clear. It acts like a very fast second pair of eyes across large volumes of content.

It doesn’t replace proper review, but it smooths the process. Fewer small issues slip through, and we spend more time focusing on the bigger things that matter.

So… Does This All Make Learning Worse?

Only if you use it badly. If you treat AI like a magic button that spits out finished content, skip the thinking, ignore the value of experience, and move straight to “publish”, then yes, you’ll end up with bland, generic learning that no one really engages with. You’ve probably already seen plenty of that!

But that’s not an AI problem. That’s a judgement problem.

Used properly, it does the opposite. It strips out the grind. The repetitive bits. The time spent wrestling something from a blank page into something usable, and it gives you that time back to do the stuff that actually matters.

• Thinking.

• Shaping.

• Challenging.

• Designing something that feels real rather than assembled.

Where We Add Value and What That Means for Clients

AI doesn’t replace what we do. It just makes it clearer where the value actually sits, and gives us more ways to deliver it.

It’s not simply about producing more content, faster. It’s about knowing:

• What actually needs to be said

• How to say it so it lands

• What the learner will care about (and what they won’t)

• Where to simplify and where to go deeper

• How to turn content into something that feels like an experience, not a document

And it’s everything around that as well.

• Working with stakeholders who all want slightly different things

• Balancing priorities

• Understanding the platform it’s going into

• Making sure it actually works in the real world, not just in theory

AI supports, but doesn’t replace. What it does give us is more flexibility in how we deliver.

In some cases, it allows us to produce content more efficiently and at lower cost, which makes it viable to create learning where previously it might not have been. In others, it gives us the space to go further. More refinement, more iteration, more detail, and ultimately higher quality where it matters most.

So it’s not a trade-off between cost and quality. It’s about having more options and choosing the right approach for the need. From a client perspective, that means better outcomes. Not just faster output, but learning that’s more considered, more relevant, and better aligned to what you’re actually trying to achieve.

Final Thought

We’re not especially interested in AI as a headline feature. For us, it’s just become part of how we work. An incredibly useful part - a transformative part, but ultimately still just a tool.

The goal hasn’t changed. It’s still about creating learning that works - AI gives us more ways to do it well.