The AI feature you’re paying extra for is about to become the baseline

Full disclosure: I work for an e-learning authoring platform. Take that however you want. 

And I’ve also spent years on the other side of the table working with L&D teams at ABN AMRO, dsm-firmenich, Signify and most recently Tesco, trying to stretch budgets that were never quite big enough. So, when I say what I’m about to say, it’s coming from both places. 

Some authoring platforms have recently started charging extra for AI-assisted authoring. The L&D community has noticed, and the reaction has been loud. I don’t think that’s just about the money. I think it’s pointing at something more fundamental, What should AI actually be doing in a learning tool in the first place? 

First, the cost story (and I’ll keep this quick) 

Stanford’s 2025 AI Index Report found that the cost to run AI at GPT-3.5 levels dropped over 280 times between late 2022 and late 2024. Hardware costs down 30% a year. Energy efficiency up 40% a year. 

Look at what Google, Microsoft, Canva and Figma have done. AI got absorbed into the product. The saving went to the customer. 

L&D software isn’t a special case. Nearly every authoring tool launched in the past two years includes AI in its base tier. McKinsey says 88% of organisations now use AI in at least one business function. At that point, “does it include AI?” isn’t a selling point anymore. It’s just a box you tick. 

So charging extra for it is basically a bet that the market won’t catch up with you. Cloud storage used to cost extra. So did collaboration features. Mobile-responsive design was once a premium add-on. Every one of those vendors eventually had to reverse course. And the reversal always cost more in credibility than the extra revenue was worth. 

My honest prediction is that every major authoring platform will include AI in its base offering within 12 to 18 months. The only question is whether this is willing or they get dragged 

But honestly, the pricing thing isn’t even the interesting part 

What I keep coming back to isn’t what AI costs. It’s what it’s for. 

There are two very different ways to use AI in learning, and they lead to completely different places. 

The first is just generating content. Give it a topic, get a course. Fast, cheap, high volume. Also, mostly not how people actually learn. A catalogue full of AI-generated modules that nobody finishes or remembers isn’t a learning strategy, it’s productivity theatre. 

The second is helping the people who already know things share what they know more effectively. Think about it. Most of the knowledge that actually matters in an organisation isn’t sitting in a document somewhere. It’s in people’s heads. The engineer who knows why the system breaks, the account manager who knows how to handle that client, the compliance person who knows where the real risk lives. The question is how you get that knowledge out of their heads and into a format that actually helps someone else. 

That’s where AI can be genuinely useful. Not instead of the author but with them. Helping them structure their thinking, tighten their explanations, get to something publish-ready without needing an instructional design degree. 

The tools you choose are quietly picking one of these two paths for you. AI that generates content sees learning as a production problem. AI that assists the author sees it as a knowledge transfer problem. They’re not the same product, even if they both say “AI” on the tin. 

What to actually do with this 

L&D budgets don’t work like engineering budgets. When your tool gets more expensive, the money comes from somewhere real as we all know. A programme gets cut, a coach doesn’t get hired, headcount doesn’t get approved. I’ve had that conversation with too many people for it to feel abstract. 

So yes, ask whether AI is included in what you’re already paying for. That’s a fair question and you deserve a straight answer. 

And ask the harder one too: what is the AI in your tools actually trying to do? Is it helping you publish faster, or helping your people learn better? Is it optimising for volume, or for something that actually sticks? 

The vendors who figured out early that AI should lower your costs, not raise them, built something real with their customers during the window that mattered. 

The ones who also figured out why AI belongs in a learning tool which is not to replace the author but to make them better  built something more durable than a pricing advantage. 

Those are the tools people will still be talking about.