How AI is changing assessment in education and workplace learning

Over the last year, AI has become almost impossible to avoid in education. Every week there seems to be another article questioning whether students are using it to complete assignments, whether qualifications still hold value or whether educators are losing the battle against technology.

It even found its way into a conversation with my mum recently.

She'd read an article claiming that student doctors were using AI to complete their degrees and that we'd soon be treated by people who had never really learned anything because "AI did everything for them." Like many headlines, it painted a worrying picture, but it also oversimplified what good education and assessment actually look like.

Doctors don't qualify because they've written an essay. They complete practical assessments, professional discussions, clinical observations and demonstrate their competence repeatedly in real situations. They take blood pressure, interpret symptoms, communicate with patients and justify the decisions they make. Those are the things that matter.

Which made me wonder whether AI is really creating a new problem, or whether it's simply exposing an old one.

Assessment has always evolved

Education has never stood still.

The way we teach has changed dramatically over the years as we've learned more about how people develop knowledge and skills. There was a time when standing at the front of a classroom and dictating notes for an hour was considered perfectly acceptable. Today, we'd question how much meaningful learning was taking place.

We've gradually moved towards more discussion, collaboration, reflection, active learning and practical application because we understand that people learn far more effectively when they engage with knowledge rather than simply receiving it.

Assessment has evolved alongside teaching.

Mathematics is a good example. There was a time when calculators weren't allowed in examinations because the expectation was that every calculation should be completed manually. Eventually, calculators became part of everyday life and assessments adapted. Nobody suggested mathematics had become less important. Instead, the focus shifted towards applying mathematical thinking rather than testing whether somebody could complete every calculation without technology.

AI feels like the next stage in that journey.

AI isn't the problem. Weak assessment is.

If a learner can produce an assignment in seconds simply by writing a good prompt into ChatGPT or Claude, it raises an interesting question.

Were we assessing their understanding, or were we assessing their ability to write an essay?

That isn't to dismiss written assignments completely. They still have an important place within education, particularly when learners are encouraged to reflect critically, draw on their own experiences and develop arguments. The challenge comes when assessment becomes predictable enough that anyone, or anything, can generate an acceptable response without demonstrating genuine understanding.

Rather than seeing AI as something that weakens qualifications, I think it's encouraging us to ask better questions about the assessments we design.

If an assessment can be completed convincingly without somebody ever applying the learning, perhaps it wasn't measuring competence as effectively as we thought.

Moving beyond the traditional assignment

One of the reasons I love teacher training is that it encourages educators to think about assessment differently.

The Level 3 Award in Education and Training, for example, certainly includes written work, but it doesn't stop there. Learners are also expected to plan learning, deliver a micro-teach, observe others teaching, reflect on their own practice and receive feedback. Those activities develop confidence and competence in a way that an essay alone never could.

The same thinking can be applied across almost any qualification or workplace training programme.

Professional discussions allow learners to explain their thinking and justify decisions in their own words. Scenario-based activities encourage them to apply knowledge rather than simply describe it. Workplace observations provide evidence of competence in real environments, while reflective accounts ask learners to consider what they've learned, what challenged them and what they would do differently next time.

Portfolios built around genuine workplace evidence remain incredibly valuable because they demonstrate progression over time rather than capturing a single snapshot of knowledge. Presentations, recorded demonstrations, role plays, case studies and project work all encourage learners to communicate, analyse and apply what they know.

None of these methods are new. What AI is doing is reminding us why they matter.

Reflection and experience can't be manufactured

One of the areas I think will become increasingly important is reflection.

AI can certainly help somebody improve the grammar of a reflective account or suggest a clearer way of structuring their writing. Used in that way, it's a fantastic tool and one I'd encourage learners to use.

What it can't genuinely do is reflect on an experience it hasn't had.

It can't explain how a difficult conversation with a learner changed somebody's teaching style. It can't describe the first time someone delivered a toolbox talk and realised halfway through that they'd lost the attention of the room. It can't reflect honestly on mistakes made during a coaching session or explain how confidence grew after delivering a successful induction.

Those experiences belong to the learner.

Good assessment should encourage people to draw on those experiences because that's where meaningful learning often takes place.

The role of educators is changing

I don't think AI makes educators less important. If anything, I think it raises the standard of what good teaching and assessment should look like.

Rather than spending time marking work that simply repeats information, educators have an opportunity to facilitate richer discussions, observe practical skills, provide coaching and create assessment opportunities that require learners to think critically and demonstrate genuine competence.

Assessment becomes less about collecting evidence for the sake of it and more about understanding what somebody can actually do. For organisations delivering workplace learning, this shift feels particularly relevant. Businesses rarely want employees who can describe a process in theory but struggle to apply it in practice. They want people who can make decisions, solve problems, communicate effectively and adapt to real situations.

Those are the skills good assessment should be measuring.

Better questions create better learning

Perhaps the biggest change AI is encouraging isn't technological at all.

For years, educators have asked learners to describe, explain, outline and identify. Those questions still have their place, but they don't always tell us whether somebody can use what they've learned.

• Maybe it's time we asked different questions instead.

• Tell me about a time you applied this in your workplace.

• Walk me through the decisions you made.

• What would you do differently next time?

• Why did you choose that approach?

• How did your actions affect the outcome?

Those questions don't simply test knowledge. They encourage reflection, judgement and application. They reveal how people think, not just what they can remember.

Teaching has always evolved

Perhaps that's why we don't see AI as something to fear.

Education has always adapted alongside society, technology and our understanding of how people learn. Every major change has challenged educators to rethink established approaches, and assessment has evolved alongside it.

AI feels like another step in that journey.

Rather than replacing good teaching or meaningful assessment, it's encouraging us to become more thoughtful about how we develop and measure competence. It pushes us to move beyond stale assignments, predictable questions and evidence gathered simply because "that's how it's always been done."

For us, that's a positive thing.

The best educators have never stood still, and neither has education itself. If AI encourages us to design more meaningful learning, create richer assessment opportunities and focus on genuine competence rather than simple recall, then perhaps it's doing exactly what good technology has always done.

Not replacing people.

Helping us become better at what we do.

About Emblem Training Solutions

At Emblem Training Solutions, we believe learning should develop genuine competence, not simply generate evidence. From bespoke digital learning and Train the Trainer programmes to our upcoming Highfield Level 3 Award in Education and Training, we help organisations create learning experiences that encourage reflection, application and lasting impact.

Find out more