An AI evaluation: What is and isn’t (yet) working in schools

Artificial intelligence (AI) has been an assistant to teachers in the classroom for the last few years, helping to streamline admin tasks, personalise learning materials and more. As with any tool, it’s always good to take stock and reflect on what is and isn’t working.

One survey suggests that 60% of teachers are using AI in some form, and many more are considering the best ways to start using the tech. There’s certainly been a lot of hype about AI in education, so it’s unsurprising that educators are keen to leverage its benefits. But it’s also important to be aware of the challenges that AI can throw up, and how to resolve them.

For your 2025 school AI evaluation, take a look at the pros and cons below.

AI in schools: What’s working

Personalised learning pathways

Perhaps the biggest impact AI has had in education is personalisation. The technology recognises how students interact with materials and what resonates with them, and then adapts accordingly. For example, some AI platforms determine what type of instructions each student follows best, and issues them in that format. Other tools calculate student response times and adjust the speed of the lessons based on that information.

This type of learning has repeatedly proven to have a positive effect on students’ academic performance, as well as boost student engagement and accessibility. The more that AI can cater to students’ needs, the easier it is for teachers to accommodate diverse learning needs and make classes more inclusive.

Administrative support

From the perspective of educators, AI has been a huge support for admin tasks like grading assignments, preparing lessons and data organisation. AI tools streamline a lot of time-consuming responsibilities that teachers have, meaning you can concentrate more on in-person guidance for students. In fact, some AI saves teachers up to 20 hours of work per week.

Plus, AI highlights valuable insights from data that inform teacher strategies. For example, if students are struggling with a particular concept, or if a specific type of content motivates students, you can then use this information to change your approach. This awareness leads to more impactful teaching and student outcomes.

Collaborative and creative projects

The coolest and most futuristic element of AI has made it possible for students to expand their imaginations and work together on artistic projects. Generative AI, virtual reality and augmented reality allow students to step into fictional worlds or scenes from the past, to learn in a much more immersive manner. Research shows too, that this type of learning helps students’ understanding, self-efficacy and motivation.

For example, VR has helped students carry out groupwork in schools in the UK. One student explores with a VR headset on and has to describe what they see for another student to transcribe. Together, the students visually, verbally and linguistically discover new worlds.

AI in schools: What isn’t working

Equity and access issues

AI resources can be expensive, and many schools (particularly in underserved areas) can’t afford to invest in the tech. Even if schools do have AI tools, the additional cost to train staff and draft ‘fair use’ policies simply may not be possible in schools with limited budgets.

There’s also an issue of accessibility within AI tools, which don’t always cater to diverse capabilities, for instance by not offering subtitles, voice command or personalisation features that make the tools and benefits available to students with disabilities.

Not to mention that the data that powers a lot of AI tools is biased, generally favouring or providing perspectives from white, able-bodied people. If the tools aren’t designed with more representative data, they can’t be inclusive for everyone.

Overreliance on AI

Nearly half of high school students say they’re already using AI, and while that means their AI literacy is developing, it could lead to them being too dependent on AI at school.

AI should be viewed as an assistive tool that builds on teachers’ expertise and makes learning more dynamic. It should not be a substitute for students’ initiative, research, or writing. If students become too reliant on AI tools, they may be at risk of losing out on skills like communication and critical thinking. Plus, it could contribute to higher rates of student cheating with AI.

Lack of AI training for teachers

AI is evolving faster than teachers can learn about it. And, for some people, AI isn’t something they can instinctively use. Most teachers will require regular training and updates about safe protocols for using AI, especially in schools.

Close to 60% of teachers say that they have had no training around AI, which creates a significant gap between the application of this technology and the benefits it can bring. It also increases digital threats and the possibility of AI misuse among students.

To streamline how teachers learn about and integrate AI tools in the classroom, schools need to implement regular, accessible AI training for educators. At the same time, the people designing AI tools need to prioritise intuitive onboarding and features that everyone can easily understand.

The (growing) age of AI

There’s no doubt that AI has made huge advances in education. It’s not only a futuristic tool, but has real value for teachers, ranging from enabling more personalised learning experiences and streamlining administrative work to enabling more collaborative, creative classrooms.

Still, like any new addition to the classroom, there are growing pains. AI has to overcome accessibility issues, overreliance and a lack of training to be more effective and fair in schools. And much like learning itself, AI’s journey will inevitably involve errors, but these are moments to review and improve, and earn better results in the end.

Further reading

Learn more about AI in schools. Read The big debate: AI and education, 7 AI tools to explore with your students in 2025 and A glossary of AI terms for teachers.