What AI is doing in education

AI tutoring tools have become significantly more capable. Khan Academy's Khanmigo, Duolingo's AI features, and various AI homework assistants can now provide personalised instruction in mathematics, reading, coding, and language learning at a level that genuinely helps students. AI tutors adapt to student pace, identify specific gaps, and provide unlimited patient repetition that human teachers cannot provide at the individual level in a classroom of thirty. In higher education, AI writing assistants, research tools, and automated feedback systems are changing how students learn and how assessors detect AI-generated work.

Administrative teaching work (lesson planning templates, report writing, marking multiple-choice assessments, progress tracking) is increasingly assisted by AI tools. This reduces the time burden on teachers but does not replace the relational work of teaching.

What AI cannot do in a classroom

Effective teaching is fundamentally human. The most important things that happen in a classroom, the moment when a student who believed they could not do mathematics succeeds at a problem and you watch their self-belief shift; the ability to read that a child is struggling at home and adjusting the day accordingly; the management of a group dynamic that involves thirty different personalities, interests, and social pressures; the mentorship relationship that shapes how a young person thinks about their own potential — none of these can be replicated by an AI tool.

Teachers also provide something that parental and social pressures have historically valued highly and that AI cannot deliver: the authority of a trusted human adult who has invested in a child's development and whose opinion matters to that child. This relationship dimension is the core of what makes teaching transformative rather than just instructional.

The future of teaching careers

Teaching is one of the most AI-resistant professions precisely because its most important functions are relational. AI tools will change what teachers spend time on: reducing time on administrative tasks, providing better data on individual student progress, and allowing differentiated practice in ways that were previously impossible at scale. But these tools augment teachers rather than replacing them. The structural driver of teaching career security in the UK is not just AI resistance but structural shortage: there are too few teachers, particularly in secondary STEM subjects and in schools in challenging areas, and this shortage is likely to persist regardless of AI tools because the relational dimension of the role requires human investment that technology cannot substitute.

Get real-time help in your next interview
Live Interview Help listens to your interview and surfaces personalised answers in real time. Free 20-minute trial on Google Meet, Teams, and Zoom.
Install Free on Chrome

Frequently asked questions

Will AI tutoring tools make it harder to justify teacher salaries?
In the short term, AI tutoring may be used by some policymakers to argue that classroom sizes can increase or that teaching assistant roles can be reduced, because AI provides individualised support that human TAs previously provided. This is a genuine risk to the teaching workforce at the margin. However, the evidence from research on educational outcomes consistently shows that human relationship quality, not information delivery, is the primary driver of educational achievement. AI tutoring that reaches students who would otherwise have no support (e.g., students in poorly resourced schools with large class sizes) is a supplement that improves outcomes, not a replacement that maintains them at lower cost.