Skip to content
Book a free discovery call

Articles

The Rise of AI in CPD - Threat or Oppurtunity?

The Rise of AI in CPD - Threat or Oppurtunity?

What happens when AI finishes CPD for you 

AI isn’t going to replace CPD. It’s going to expose which CPD was never doing anything in the first place and quietly recreate the same problem if associations aren’t careful. 

A member recently told us they’d finished their annual CPD requirement in an afternoon. Not by attending a single session, but by asking an AI tool to summarise the reading, generate the reflection notes, and tick the box. Nobody checked; nobody could have. The requirement was met , and yet nothing was learned. 

That’s the uncomfortable part of this conversation: AI can make learning faster, more personalised and more accessible than ever. It can also make it easier to look like learning happened when it didn’t. The tools aren’t the problem; the absence of accountability around them is. 

What AI actually changes about learning 

Used well, AI changes CPD in ways that were genuinely difficult before. A member struggling with one specific part of their practice can get a personalised pathway instead of sitting through a module built for the average learner, because there is no average learner. Instant Q&A means a question gets answered the moment it comes up, while the problem is still fresh. Adaptive scenarios can target someone’s actual weak spots rather than walking everyone through the same fixed curriculum. 

But responsiveness isn’t the same as depth. A personalised pathway still needs the person to sit with the material, wrestle with it, get something wrong and understand why. An instant answer only helps if the member absorbs it, rather than using the AI as a shortcut past the thinking altogether. 

So the real question isn’t “should we allow AI in CPD?” It’s “are we building learning experiences that AI can enhance, or box-ticking exercises that AI can simply finish for people?” The first is the opportunity. The second is the risk. Most associations haven’t yet decided which one they’re running. 

The risks nobody’s pricing in 

Verification: If a member submits a reflective learning log, how does anyone know whether they wrote it or an AI produced it in ninety seconds from a two-line prompt? Right now the honest answer is: nobody knows. The log looks the same either way, ticks the same boxes, satisfies the same requirement. 

Devaluation: A huge amount of CPD content has always been generic the same foundational material delivered the same way to everyone. AI can now reproduce that in seconds. If a session can be replaced by a well-worded prompt, it was never a learning experience; it was information delivery wearing the costume of one. 

Gaming: Members meeting point requirements through AI-assisted box-ticking not out of malice, but because the system was built to reward completion, not understanding. If the requirement is “prove you did the thing,” and AI can prove it convincingly without the thing happening, most people under time pressure will take the shortcut. 

None of these are really about AI. They’re about accountability structures that were already thin, now tested by a tool efficient enough to expose exactly how thin. 

Where AI actually earns its place 

None of this means AI has no place in CPD. It means its place needs to be chosen deliberately, not backed into by default. Used well, AI can act as a tutor between formal sessions, handle the administrative weight of reflective CPD, and match members to the learning that’s genuinely relevant to their role and gaps. 

Each of these has one thing in common: AI does the work around the learning, not the learning itself. That’s the distinction that matters: 

  • AI organising evidence of reflection = opportunity. AI generating the reflection = risk. 

  • AI matching a member to the right course = opportunity. AI completing the course = risk. 

The line isn’t complicated, but it does require associations to draw it rather than letting members find it for themselves through trial and error. 

Why the rules need to come before the questions 

Most members using AI for CPD aren’t waiting for permission. They’re already doing it quietly, because no one’s told them not to and no policy says otherwise. That’s the actual state of play, whether associations have acknowledged it or not. 

Regulators and associations will need to define what “AI-assisted” CPD even means before they can enforce anything. Is it acceptable for AI to help organise a reflection but not write it? To explain a concept but not generate the proof of engagement? These aren’t hypothetical questions for some future date, members are answering them for themselves, one submission at a time, in the absence of any guidance. 

The associations that define this now, clearly, before a scandal forces the issue are the ones that will look credible later. First-mover advantage here isn’t about being innovative, it’s about not being the association still writing its AI policy after everyone already knows the old system had a hole in it. 

What associations should do next 

If the distinction above is the principle, here’s how to put it into practice in roughly the order it pays to tackle it. 

  • Publish a plain-English AI-use position: One page. State what AI may assist with (organising evidence, explaining concepts, suggesting pathways) and what members must still do themselves (the reflection, the judgement, the applied thinking). Members will follow a clear line far more readily than a vague one. 

  • Redesign the requirement around evidence of thinking, not proof of completion; Replace “summarise this reading” with “describe a decision you changed as a result, and why.” Prompts that ask for applied, situated judgement are the ones a generic AI answer can’t fake. 

  • Build verification into the workflow, not the honour system; Timestamps, reflection captured at the point of learning, light supervisor or peer sign-off for higher-stakes CPD. The goal isn’t surveillance — it’s making the genuine route easier than the shortcut. 

  • Let AI carry the admin, so members spend their time on the learning; Auto-gather and organise evidence, chase the paperwork, structure the log. Reserve human effort for the parts that actually build competence. 

  • Review your content library honestly: If a module can be replaced by a well-worded prompt, retire or rebuild it. What survives should be the material that needs a human to engage with it. 

A CPD platform makes the difference between a policy on paper and one that holds. They are built to support capturing reflection at the point of learning, organising evidence automatically, and giving associations an audit trail that stands up to scrutiny. The technology doesn’t replace the judgement about where AI belongs. It enforces the line once you’ve drawn it. 

Why the real threat was never AI 

People have always found faster ways to learn, and most are genuine improvements. The real threat is associations pretending AI changes nothing, while members already use it every day to get through requirements built for a world before it existed. 

CPD was supposed to prove that learning happened. AI hasn’t broken that promise. It’s just made it a lot easier to notice how many CPD programmes were never really keeping it.