By The HR Horizon | Learning & Development | Future of Work
Reading time: approximately 9 minutes
Learning and development (L&D) has always been one of the clearest signals that employees look for when deciding whether to stay with an employer or move on. What’s changed is what that development now needs to include. It’s no longer enough to build capability in the skills that your business has always needed — you also need to build capability in the tools that are actively reshaping how that work gets done. Right now, that means AI.
For many SMEs and startups, AI adoption has happened by accident rather than by design. Someone on the team started using an AI writing tool. Someone else found a way to speed up data analysis. A manager quietly started using AI to draft first-cut job descriptions or meeting summaries. None of it was planned, none of it was shared across the business, and none of it was built into how people are developed. That’s the gap that this guide is about closing.
The good news is that an AI-ready L&D strategy doesn’t require a large budget or a specialist training department. It requires the same thing that effective L&D always has: intention. You need to know what capability your people actually need, build it deliberately, and make sure it compounds across the team rather than living in the heads of a few early adopters.
“AI upskilling isn’t a separate initiative bolted onto your L&D plan — it’s the lens your entire L&D plan now needs to be built through.”
Start With the Right Question
Most small businesses approach L&D by asking, “What training should we do?” That question leads to generic courses and low completion rates. The better question is: “What capability does our business need over the next 12 to 24 months that we don’t currently have — and who needs to build it?”
With AI in the picture, that question now has two layers. The first is the capability your business has always needed — the sales skills, the technical knowledge, the leadership behaviours. The second is newer: which of those capabilities can be meaningfully accelerated, augmented, or extended by AI, and does your team know how to do that yet? In our experience, most SMEs haven’t asked that second question at all. They’re either ignoring AI, banning it out of caution, or leaving employees to figure it out alone — none of which builds real organisational capability.
A Capability Gap Analysis That Includes AI
Run the same structured process you’d use for any capability gap — but widen the lens:
- Define the key goals your business needs to achieve in the next year.
- Identify the skills, knowledge, and behaviours required to achieve them — including where AI tools could reasonably speed up or improve that work.
- Assess your current team honestly: who is already using AI well, who isn’t using it at all, and who is using it in ways that create risk (unreviewed AI-generated content, sensitive data pasted into public tools, and so on)?
- Identify the gaps — both the traditional skills gaps and the AI-fluency gaps.
- Prioritise: which gaps, if closed, would free up the most time, reduce the most risk, or unlock the most capacity?
The output is your L&D priority list. It should name AI fluency as one of the capabilities you’re building — not treat it as a side conversation that happens informally.
Action Tip: Run a 10-minute AI-use audit with each team: ask what tools people are already using, for what, and how confident they feel using them responsibly. You’ll usually find far more AI use already happening than leadership assumes — and several risks nobody’s flagged yet.
The 70:20:10 Model, Applied to AI Skills
The 70:20:10 model is one of the most useful frameworks in workplace learning, and it happens to be perfectly suited to building AI capability on a small budget. It holds that:
- 70% of learning happens through on-the-job experience — real tasks, real problems, real repetition.
- 20% happens through learning from others — coaching, mentoring, peer exchange, and observation.
- 10% happens through formal learning — courses, workshops, and structured training.
Applied to AI, this is genuinely good news for resource-constrained businesses. You don’t need to send your whole team on an expensive AI course to build real capability. However, you do need to be intentional about how you leverage the 70% and 20% that’s already available to you, across your business, at little to no direct cost.
The 70%: Build AI Fluency Into the Work Itself
This is where most of the capability-building should happen. Give people real tasks and explicitly ask them to try an AI-assisted approach alongside their usual method be it drafting a first version of a report, summarising a long document, generating options for a proposal, or cleaning up messy data. The point isn’t to hand the task entirely to AI. It’s to build the judgement to know when AI genuinely helps, when it doesn’t, and how to check and improve what it produces.
The 20%: Learning From Each Other
Some of your team are already ahead on this. Structured peer learning — which could be a simple 20-minute session where someone walks the team through the AI workflow that they’ve built for themselves — costs nothing but time, and it spreads capability far faster than any course. This is also where a lot of the risk-reduction happens: experienced users can model good judgement about what should and shouldn’t go into an AI tool, and what still needs a human check before it goes out the door.
The 10%: Formal Learning, Used Well
Formal AI training still has a place — particularly for building a shared baseline of understanding (what these tools actually do, where they get things wrong, what “responsible use” means in practice) and for role-specific deep dives where the 70/20 approach won’t get people far enough on their own. Online courses built specifically for business and HR contexts — like the ones we offer at The HR Horizon — are an efficient way to cover this 10% without the cost of custom in-house training design.
“The organisations that treat AI as a threat to manage will lose talent to the ones that treat it as a capability to build.”
High-Impact, Low-Cost Ways to Build AI Capability
1. Stretch Assignments — With AI Built In
Give a high-potential employee a project outside their usual scope, and explicitly ask them to identify where AI tools could help them move faster or think more broadly. This does double duty: it’s a classic stretch assignment, and it builds AI judgement in a real, high-stakes context rather than a sandbox exercise nobody takes seriously.
2. Internal Knowledge Sharing on AI Use Cases
Your team collectively knows more about useful AI applications than any single manager does. A monthly “lunch and learn” where someone shares a workflow they’ve built — how they’re using AI to draft first-pass customer responses, summarise meeting notes, or speed up a repetitive admin task — costs only meeting time and builds capability across the whole group at once.
3. External Online Learning
The quality and accessibility of online learning on AI has grown quickly, and much of it is inexpensive relative to traditional training. Platforms like The HR Horizon’s courses — built specifically for SME leaders and HR professionals across the Caribbean and beyond — cover both the practical mechanics of using AI tools responsibly and the broader future-of-work context leaders need to plan around.
4. Mentoring Programmes
Pair employees who’ve built genuine AI fluency with those who haven’t. This transfers not just tool knowledge but judgement — when to trust an AI output, when to push back on it, and when a task simply shouldn’t go through AI at all (sensitive employee data and legally significant documents being the obvious examples). For SMEs in the Caribbean, cross-company mentoring arrangements with non-competing businesses can widen this pool further.
5. Action Learning Sets
Small groups meeting regularly to work through real challenges together are a natural fit for AI adoption questions: “How do we use this without exposing client data?” “Where is this actually saving us time versus just moving the work around?” “What do we do when the output is confidently wrong?” These are exactly the questions your team needs to work through together, out loud, rather than each person guessing alone.
6. Leadership Coaching for the AI Era
Leaders now need to make decisions about tooling, data governance, and workforce planning that didn’t exist on this scale a few years ago. A small number of executive coaching sessions focused specifically on leading through this shift — how to set expectations, how to manage anxiety about job security, how to model good use rather than either blind enthusiasm or blanket restriction — can produce an outsised return relative to the time invested. This is a core part of the leadership coaching we do at The HR Horizon.
7. Conference and Event Attendance
Sending a rotating group of employees to industry events, AI-focused masterclasses, or professional association sessions brings outside perspective back into the business and signals that development — including AI development — is genuinely invested in, not just tolerated.
Action Tip: Pick one low-cost approach from this list and pilot it with a single team this month. Don’t try to roll out all seven at once — a focused pilot that actually gets used beats a comprehensive plan that stalls at the launch stage.
What AI Upskilling Actually Looks Like at Different Levels
“AI upskilling” means something different depending on where someone sits in the business, and a flat, one-size-fits-all training session tends to miss most of the team.
Frontline and Operational Staff
For most operational roles, the priority is practical: using AI to handle repetitive drafting, summarising, and data-tidying tasks so more time goes to the parts of the job that need a human judgement call. The goal is comfort and competence with a small number of tools used well — not broad theoretical knowledge.
Managers and Team Leads
Managers need a layer above tool competence: judgement about when AI-assisted work from their team is good enough to go out, and when it needs a closer look. They also need to be able to answer their team’s questions about what’s appropriate to use AI for — which means they need clearer guidance from leadership than “use your best judgement.”
Senior Leadership
At this level, AI upskilling is less about using specific tools and more about strategic fluency: understanding where AI genuinely changes the business’s competitive position, where it introduces real risk (data protection, intellectual property, employment law questions around AI-assisted decisions), and how workforce planning needs to adapt as a result.
“If your AI training only reaches one level of the business, it will build a capability gap between levels instead of closing the one you started with.”
Building Your L&D Plan: A Practical Template
For each priority capability gap you’ve identified — including AI-specific ones — work through the same five elements:
- Name the capability and why it matters for the business.
- Identify who needs to develop it — individuals, a specific team, or the whole organisation.
- Select the most appropriate, cost-effective development approach from the 70:20:10 mix above.
- Define a timeline and any associated costs.
- Agree how you’ll know the development has worked — what will you actually observe changing?
A simple annual plan covering these five elements for your top five to ten priority capabilities — with at least one or two explicitly AI-related — is enough to get started. It doesn’t need to be elaborate. It needs to get used.
Making AI Fluency Part of the Culture
Budget and activity alone won’t build a learning culture, and this is especially true with AI, where anxiety often runs ahead of understanding. If leaders are quietly worried AI will make their own role redundant, that worry filters down — whether or not anyone says it out loud. If senior leaders visibly engage with AI themselves, talk openly about what they’re learning and where they’ve gotten it wrong, and treat questions about job security honestly rather than dismissing them, that signals safety. If they don’t, no amount of formal training will compensate.
Practical ways to build this into a small business:
- Make AI use and its limits a standing conversation in team meetings, not a one-off announcement.
- Celebrate good examples of AI-assisted work — and be equally open about the times it went wrong and what was learned.
- Create real psychological safety around AI use: people should be able to admit they don’t know how to use a tool, or that they made a mistake with one, without it counting against them.
- Set clear, written boundaries on what should never go into an AI tool — client data, personal employee information, anything commercially sensitive — so people aren’t left guessing.
- Address job-security anxiety directly rather than hoping it goes away on its own. Silence reads as confirmation of the worst fear, not reassurance.
Measuring the Return on Your AI Upskilling Investment
L&D budgets get cut when their impact isn’t measured, and AI upskilling is no exception. Track a mix of the practical and the cultural:
- Application: are people actually using what they’ve learned in their day-to-day work? Ask managers to observe and report specific examples.
- Time and capacity: is AI-assisted work freeing up meaningful time, and is that time being redirected to higher-value tasks?
- Quality and risk: is AI-assisted output being properly checked before it goes out, and are there any near-misses worth learning from?
- Confidence and engagement: has your engagement survey data shifted on items related to development, tools, or confidence in the role?
- Retention: are employees who receive structured development — AI-related or otherwise — staying longer than those who don’t?
Even simple, informal tracking against these gives you the evidence to justify continued investment and to refine your approach as the tools themselves keep changing.
Final Thoughts
The businesses that get the most value from AI over the next few years won’t necessarily be the ones with the biggest technology budgets. They’ll be the ones whose people know how to use it well — with judgement, with appropriate caution, and with genuine capability rather than guesswork. That comes down to L&D, and L&D comes down to intention.
Start with your capability gaps, including the AI-specific ones. Build a focused, blended approach using the resources you already have. Address the anxiety honestly instead of ignoring it. And make sure the capability you build spreads across the team — not just the two or three people who taught themselves first.
EXPLORE THE HR HORIZON COURSES | Online, self-paced, built to help businesses own the horizon.
About The HR Horizon
The HR Horizon is a Caribbean-based HR consultancy and learning platform helping SMEs, startups, and emerging leaders build future-ready organisations. We offer HR consulting, executive coaching, online courses, managed HR services and a library of ready-to-use HR templates and policies designed specifically for businesses across the Caribbean and beyond. Visit us at thehrhorizon.com or email hello@thehrhorizon.com.



