Organisations everywhere are investing in AI training.
Employees attend a workshop, learn how to write a prompt, experiment with a few tools and leave excited about what AI might be able to do.
Then they return to work.
Some continue experimenting. Some use AI occasionally. Others are unsure where they are allowed to use it, how it applies to their role or whether they are using it effectively.
Within a few months, organisations finds themselves with pockets of enthusiastic AI users, employees who barely use it at all and teams developing their own practices. Net result, an consistent approach.
The problem is not necessarily the training. It’s treating AI training as a one-off event.
Building genuine AI capability requires a more structured approach. Organisations need to consider how people progress from understanding AI to applying it effectively in their roles, embedding it into workflows and using it responsibly over time.
That is the thinking behind my AI Capability Development Pathway.
Stage 1: Develop leaders first
Successful AI adoption starts with leadership.
Before organisations expect employees to change the way they work, leaders need to understand what AI can and cannot do, where the opportunities lie and what responsible use looks like.
It does not mean every leader needs to become an AI expert. They do, however, need enough knowledge to ask the right questions, identify opportunities, manage risks and support their teams through change.
AI Leadership training should go beyond learning how to use an AI tool to explore:
- responsible AI use
- critical thinking
- governance
- workforce implications
- how to lead teams through technological change.
When leaders understand AI, they are better equipped to champion its use with clarity and confidence.
Stage 2: Build foundational AI capability across the workforce
The next step is to give employees the practical skills they need to use AI effectively in their everyday work.
For many employees, one of the most immediate applications is business writing.
Emails. Reports. Briefing notes. Summaries. Correspondence. Meeting notes. Proposals.
AI can help employees improve and streamline all these tasks, but only when they know how to communicate effectively with the technology. It requires more than typing a basic instruction into a chatbot. Employees need to understand how to provide context, define an audience, specify a purpose, refine outputs, question results and verify information.
They also need to understand that AI does not replace their professional judgement.
A foundational program creates a common level of AI literacy across the organisation and gives employees the confidence to begin using AI in practical, relevant ways.
Stage 3: Move from generic training to team-specific workflows
It is at this stage that AI training becomes particularly powerful. Different teams do different work meaning a finance team, HR department, clinical administration team, communications unit and executive office will not use AI in exactly the same way.
Their responsibilities are different. Their risks are different. Their information is different. Their workflows are different.
So why would their AI training always be the same?
Once employees master foundational AI skills, the next stage should focus on the work they actually do. In a facilitated, hands-on workshop, teams identify:
- Which tasks consume the most time?
- Which processes are repetitive?
- Where are we duplicating effort?
- What information do people regularly need to summarise, restructure or communicate?
- Which workflows involve unnecessary steps?
- Where could an LLM assist without compromising professional judgement, privacy or accountability?
The objective is not simply to automate existing tasks. Sometimes the more important question is whether the existing process needs to be redesigned at all.
A cumbersome workflow does not necessarily become a good workflow simply because AI makes it faster.
The greatest gains may come from combining AI capability with process improvement: examining how work is currently done and identifying a better way to do it.
Stage 4: Turn organisational AI policy into practical team guidance
Most organisations recognise the need for an overarching AI policy. But a high-level policy rarely answers practical question employees encounter in their day-to-day roles. Hence the need for purpose-built AI policies unique to each team that outlines:
- What does responsible AI use look like for this particular team?
- Which tasks are appropriate for AI assistance?
- What information can and cannot be entered into an approved tool?
- Who is responsible for checking an AI-generated output?
- When is human review required?
- Which tasks should never be delegated to AI?
Yes, there may be some crossover, but in practise different teams need different answers. The next stage of AI capability development should therefore involve translating organisation-wide policy into clear, team-specific practices.
It does not mean every team creates its own rules independently. But they create guidelines that align with the organisation’s broader governance framework while reflecting the realities of their specific roles and responsibilities.
The result is a much clearer understanding of not simply whether employees can use AI, but how, when and why they should use it.
Stage 5: Review, refine and evolve
AI capability development does not have a finish line. The technology is simply evolving too quickly.
Tools change. Organisational policies develop. Employees discover new use cases. Unforeseen issues emerge. Some AI applications may prove invaluable, while others turn out to be less useful than expected.
Hence, organisations need an ongoing tune-ups for reviewing AI use to assess:
- What is working?
- What is saving time?
- What has not worked as expected?
- What new risks or concerns have emerged?
- What successful practices can be shared with other teams?
- What new tools or capabilities should we explore?
It transforms AI adoption into a cycle of continuous improvement rather than a one-off technology rollout.
From AI training to AI capability
The goal should not be to train employees to use a particular AI tool as they will continually change.
The more valuable investment is developing people who know how to think critically about AI, identify appropriate use cases, redesign inefficient workflows, evaluate outputs and adapt as the technology evolves.
That requires a progression:
Develop leaders → Build foundational capability → Improve team workflows → Embed governance → Review and evolve
A single AI workshop can be an excellent starting point. But it should be exactly that.
A starting point.
The organisations that gain the greatest long-term value from AI are likely to be those that stop thinking about AI training as an event and start treating AI capability as something that must be developed, embedded and continuously improved.
Because successful AI adoption is not about teaching people how to use AI once.
It is about helping them become better at using it over time.