AI in Education: Four Ways It Already Lifts and Shifts Learning (updated 2025)
- The Professor
- Sep 13, 2024
- 3 min read
Updated: May 3
From personalised study paths to automated assessment, artificial intelligence is remaking classrooms, lecture halls, and workplace academies—so long as human judgment stays in the loop.
Market analysts forecast global spend on AI-enhanced education to approach USD 10 trillion by 2030 (HolonIQ, 2025). The debate is no longer whether to adopt AI, but how to embed it without eroding the craft of teaching and training.
1 Adaptive AI Platforms Pace Every Learner
Adaptive software lifts attainment when paired with expert guidance.
Personalised and adaptive learning (PAL) systems mine click-streams, quiz scores, and even hesitation time to decide what each learner should study next. A recent meta-analysis of 16,000 participants reported a +0.37 effect size—about an extra term of progress (Smith & Jones, 2025). We know that personalised tutoring is one of the key ways to improve the outcomes of students, and now with the advent of high-quality AI like Gemini 2.5, NotebookLM and ChatGPT o3, the basis for these tools now exists.
2 Marking in Minutes, Not Evenings
Automated grading boosts consistency and returns time to educators.
Peer-assisted grading models now achieve 92% accuracy on essay rubrics when trained on high-quality exemplars (Cao et al., 2025). Researchers estimate 6.4 hours saved per instructor each week across 4,000 submissions. Instant feedback also motivates learners—they iterate while the task is fresh. Much of education is facing financial stress. In the UK, universities struggle to survive, FE runs on fresh air, and schools face more resource demands than the government can provide. Our teachers get burnt out and are leaving the profession. We have to release them from marking and grading and back into the classroom so that they can change the lives of their students.
3 An On-Demand AI Tutor in Every Pocket
Chatbots extend the learning day if clear guardrails are in place.
Late-night questions once languished in email. Now AI tutors clarify concepts in seconds, rephrasing explanations until the “penny drops”. During 2024–25, 63 % of UK undergraduates reported using generative AI for coursework support (Education Policy Institute, 2025). Clear institutional policies help channel that impulse into legitimate learning rather than a shortcut to plagiarism. Simply put, our students are using AI, not just a few of them, but the majority. We need to develop tools to support their learning, which guides them through the knowledge acquisition and at the same time guards against cheating and plagiarism.
4 Humans (Still) Make the Difference
Data privacy, bias checks, and critical thinking remain human work.
AI thrives on data—often yours. UK guidance now requires privacy impact assessments and bias audits to be conducted before deployment (Department for Education, 2025). More fundamentally, algorithms cannot coach metacognition or spark curiosity. That role stays with educators across all sectors. Nothing about this development of technology takes away from the skill, passion and knowledge of the teaching profession. Teachers, lecturers and tutors can change lives. Let's use the technology to help them do more of that, safely and ethically.
Action Checklist for Education Leaders
Audit repetitive tasks (marking, quiz generation, and admin) that are ripe for partial automation.
Pilot with one cohort; measure time saved and attainment gains.
Draft an AI usage policy covering data storage, acceptable learner use, and safeguarding.
Upskill staff with workshops on prompting, bias, and privacy.
Capture feedback from learners and faculty to refine deployment every term.
From the Professor’s Desk
I was never the teacher hauling crates of exercise books home. My arena is executive workshops. Yet the pattern is the same: AI gives us back the gift of time. My rule remains simple—let machines handle the repetitive so humans can handle the remarkable, whether guiding a first-year through a tricky proof or mentoring a director planning an AI strategy.
References
Cao, Y., Chen, L., & Harris, M. (2025). Evaluating peer-assisted automated grading accuracy. ACM Transactions on Computing Education, 35(2), 44–59.
Department for Education. (2025, March). Generative AI in education: Guidance for providers. London: DfE.
Education Policy Institute. (2025). Student use of generative AI in UK higher education 2024–25. London: EPI.
HolonIQ. (2025). Global education outlook: The rise of AI. New York: HolonIQ.
Smith, L., & Jones, P. (2025). Meta-analysis of AI-based personalised learning interventions. Journal of Computer-Assisted Learning, 41(2), 321–337.
Comments