AI Offers Opportunities But Education Must Adapt

We’re at a turning point in education – and AI is the driving force behind it.

Coleg y Cymoedd is a further education institution based in the South Wales Valleys, with four campuses and around 10,000 learners. We support students across a wide range of levels, from entry right through to degree. And like many colleges, we’re exploring how AI can help us do that better.

For me, there are three main aspects to this. First, the potential for AI to deliver real value for learners. Second, how it can improve operational efficiency across the college. And third, the question of future skills – both for our students and for the staff who support them.

On the learner side, AI presents a genuine opportunity to personalise and accelerate the learning journey. Adaptive tools can help students engage with content in a way that fits their needs, supporting them to build understanding and confidence over time. We’ve started to experiment with this in small but meaningful ways, identifying areas where AI can make an immediate difference, particularly where we can demonstrate clear value.

At an institutional level, there’s also a huge opportunity to automate the administrative side of things. In the public sector, efficiency matters. And AI – particularly some of the earlier-generation tools that pre-date generative models – can help us reduce workload and free up time for teaching and learning. That’s work we’re already doing.

But perhaps the biggest question is how we equip our students for a future that none of us can fully predict. The world of work is changing rapidly. AI will have a hand in reshaping the jobs people do, the skills they need, and how we prepare for both. That requires us to think differently about the role of education itself and how we ensure it remains relevant.

There’s real optimism here, especially when it comes to equity. We’re already seeing research that suggests AI can benefit lower-level learners more significantly than their higher-level peers. If used well, AI could help level the playing field. But that’s only possible if we make sure that access isn’t a barrier.

The digital divide remains a serious concern. We know that many learners still face obstacles when it comes to connectivity or simply having access to a device. Thanks to Welsh Government support, we’ve been able to provide devices and internet access where needed since the pandemic, but we can’t assume that infrastructure alone is enough. We must continue to ensure that all learners can engage meaningfully with these tools and that means thinking about cost, training and digital confidence too.

The good news is that there are increasingly accessible versions of AI tools available. As a Microsoft college, we’ve been working closely with Microsoft to roll out CoPilot across the organisation. With data protection protocols in place for both staff and learners, we’re seeing how AI can be used safely and securely. Free-to-use options, including browser-based access, also help us widen participation. But we have to stay focused on inclusion – because if we don’t, we risk deepening the digital divide.

Digital transformation is not about tech for tech’s sake. It’s about helping people do what they do, but better. Bringing staff and students with us on this journey is essential. We’re building awareness, sharing best practice and developing relationships with external partners to help us grow our understanding and application of AI in a way that’s relevant and useful.

The human aspect of teaching has never been more important. Great teachers have always known how to build relationships and identify what individual learners need. What AI can do is support that process, giving us new ways to spot misconceptions and intervene early. It’s not about replacing teachers. It’s about equipping them with better tools to do the job.

That’s especially important when you consider the scale we work at in further education. Teachers have large cohorts to support. Anything that helps us understand each learner’s journey more clearly is going to be valuable.

But that also brings us to one of the biggest challenges we now face: assessment.

AI has made traditional forms of assessment harder to justify. If a learner can generate a passable answer with a few clicks, we have to ask ourselves what the point of that task really was. We can’t police this away – and we shouldn’t try. Instead, we need to rethink what assessment is for.

For me, that means a fundamental shift towards thinking of learning as a process rather than an outcome. Assessment should reflect that. It should give us insight into how a student is learning and not just whether they’ve produced the right answer. And that calls for radical redesign, from awarding bodies to specifications and teaching practices. It’s not easy, but it is necessary.

Ultimately, this is about ensuring that education continues to serve the needs of learners and the wider world. AI is not a shortcut. It’s a catalyst. If we get this right, we have a chance to build something that’s more responsive, more inclusive and more effective than what came before. But it will take collaboration, ambition and a willingness to ask hard questions.

The opportunity is there. So is the responsibility. Now we need to meet both.

Rory Meredith talks about this and more in the DeepLearn HS podcast episode AI and Education – Shaping the Future of Learning with AI. Listen to the podcast here.