AI Is Table Stakes in Learning. Adoption Is Another Story..

Everyone's talking about AI. LinkedIn feeds are stuffed with it. Every RFP we see in digital learning mentions it. If you're pitching a platform, a course, or a training programme in 2026 and can't answer the AI question, you're out of the room before the coffee arrives.

That's the headline. What's underneath is messier, and worth a closer look if you run L&D.

The expectation: AI everywhere, all at once

L&D buyers assume AI is baked into every proposal. Industry commentary this year calls AI-driven personalisation “table stakes” for 2026, and the numbers back it up. Synthesia's 2026 L&D report found that 88% of L&D teams already see value through time saved on content creation. 40% report better learner satisfaction in programmes where AI features are live.

On paper, we've arrived. In boardrooms and vendor pitches, anything less than an AI story sounds dated.

The reality: a big adoption gap

Dig a bit deeper, though, and the picture changes.

Gallup's workplace data suggests only around 10% of the US workforce uses AI daily at work. Weekly use sits higher, but it's concentrated in a handful of roles. Personal use, meanwhile, keeps climbing. People are using ChatGPT to draft birthday messages, plan holidays and troubleshoot plumbing at weekends, then showing up to jobs where the official AI strategy is a policy memo.

BCG's AI at Work 2025 research makes it sharper: investment is up, outcomes aren't keeping pace. 59% of companies invest over $1 million a year in AI, but only 29% report meaningful returns. About three quarters of executives quietly admit their AI strategy is “more for show” than real guidance. And more than half of the global workforce has had no recent AI training.

The gap between “we've deployed AI” and “our people actually use it well” is wider than most vendor decks let on.

Why this matters for L&D

Here's the part every learning leader should pay attention to.

Your workforce is already using AI. They're using it quietly, on personal accounts, without training, guardrails or context. That's the “shadow AI” problem. It looks like a security issue, but it's a learning one. When people teach themselves by trial and error, they pick up bad habits, miss the better use cases, and build workarounds that won't scale.

So yes, AI is table stakes. But the question has shifted. The one to ask now isn't whether to adopt it. It's whether your people are getting real support to use it well.

What actually moves the needle

If you're an L&D leader wondering where to focus, three things beat buying another tool:

■ Build AI literacy across roles. DataCamp's 2026 research shows organisations with mature, company-wide AI literacy programmes are roughly twice as likely to report strong AI returns. Literacy is what drives the ROI, not the tools themselves.

■ Embed AI in the flow of learning. An AI tutor that lives inside a course and remembers a learner's gaps is useful. A chatbot bolted onto a homepage is a gimmick. The difference comes down to design.

■ Keep the human loop intact. AI can scale support. It can't replace coaching, reflection or accountability. The best programmes this year blend AI with real mentorship. The weakest assume the tech does the job on its own.

A quick word on where we sit

At Capytech, we've been building this into our work with clients across the region for a while. AI coach features, AI tutor modules, AI-driven feedback, adaptive assessments, smart narration. Not because it reads well in a pitch, but because clients in banking, energy, government and retail are asking the same thing you probably are: how do we make AI actually useful, day to day?

If that's the conversation on your desk this quarter, you're asking the right question.

Over the next year or two, the companies that get real value out of AI won't be the ones with the longest feature list. They'll be the ones whose people actually know how to use what they've got.

DAMIAN HEHIR
Co-Founder & CEO Capytech

linkedin.com/in/damianhehir

www.capytech.com