VeltrixVeltrix.
← All articles
44 / 62March 31, 2026

AI in Education: How AI Is Changing Teaching and Learning in 2026

AI in education — personalised tutoring, automated grading, accessibility tools, and the academic integrity debate. With real data from Khan Academy and other deployments.

Industry / Education

AI in Education

Personalised tutoring, automated assessment, and the cheating debate — how AI is transforming education and what teachers, students, and institutions need to know.

$40B
AI in education market by 2030 — growing at 36.2% CAGR as personalised learning tools gain adoption globally [Grand View Research]
2x
Learning speed improvement for students using Khanmigo (Khan Academy's AI tutor) in randomised trials vs traditional instruction [Khan Academy]
89%
Of students globally who now have access to AI tools like ChatGPT — creating an academic integrity crisis schools are still figuring out [EDUCAUSE]

AI in education sits at an uncomfortable intersection. On one side: genuine evidence that AI tutors can personalise learning at scale, providing every student with the kind of one-on-one attention previously available only to the wealthy. On the other: AI has made academic dishonesty trivially easy, and most institutions haven't adapted their assessment methods to account for it.

AI tutoring
One-on-one AI tutors that adapt to a student's pace, identify misconceptions, and provide unlimited practice. Khanmigo (Khan Academy), Socratic (Google), and Duolingo's AI tutor demonstrate real learning gains.
Khanmigo, Duolingo, Brilliant, Coursera AI
Automated grading
AI grades multiple choice, short answer, and increasingly essays — freeing teachers from high-volume assessment tasks. Turnitin and Gradescope use AI to provide consistent grading at scale.
Turnitin, Gradescope, EssayGrader, Writable
Personalised curricula
Adaptive learning platforms adjust difficulty, pacing, and content format based on individual student performance data. Carnegie Learning's AI tutor has been shown to improve math outcomes by 23% in controlled studies.
Carnegie Learning, DreamBox, IXL, Knewton
AI writing assistance
Students use AI for feedback, revision suggestions, and research assistance. The line between assistance and generation is contested — schools are developing policies that acknowledge AI exists while still assessing genuine learning.
ChatGPT, Claude, Grammarly, Hemingway App
Teacher productivity tools
AI generates lesson plans, differentiated worksheets, assessment rubrics, and parent communication templates. Saves teachers 5-10 hours per week on administrative tasks — redirecting time to classroom presence.
MagicSchool AI, Eduaide.ai, SchoolAI, Diffit
Language learning AI
AI conversation partners for language practice — available 24/7, infinitely patient, adaptive to level. Duolingo's AI features generated 20% higher daily active usage compared to non-AI exercises.
Duolingo, Babbel, Pimsleur, Speak app
The case for AI in education
  • Personalised tutoring available to all students, not just those whose parents can afford private tutors
  • Identifies learning gaps and misconceptions early, before they compound
  • Frees teachers from repetitive marking to focus on relationships and complex teaching
  • Students learn to work with AI tools they'll use throughout their careers
  • 24/7 learning support for students in underserved communities or with caring responsibilities
The concerns
  • AI makes it trivially easy to produce essays without genuine learning
  • Students who rely on AI bypass the productive struggle that builds deep understanding
  • AI-generated feedback may not match the nuance of experienced teacher feedback
  • Algorithmic bias in adaptive learning systems may disadvantage certain student groups
  • Over-reliance on AI assistance may leave graduates unprepared for contexts where AI isn't available
The productive middle ground
The most effective educational institutions aren't banning AI or ignoring it — they're redesigning assessment. Portfolio-based assessment, oral examinations, in-class work, and tasks that require demonstrating process (not just output) are all robust to AI assistance. Teaching students to use AI as a tool while maintaining genuine skill development is the challenge — the same challenge that emerged with calculators, which were ultimately integrated rather than banned.
Is using AI to help with homework cheating?
It depends on the assignment's intent and your institution's policy. Using AI to understand a concept you're stuck on: generally fine. Using AI to write an essay that's supposed to demonstrate your own writing and thinking: academically dishonest. The key question is: are you learning, or are you outsourcing learning? The former has long-term value; the latter doesn't.
What's the best AI tool for students?
For research and understanding complex topics: Perplexity (cites sources) or Claude (good at long documents and nuanced explanations). For maths: Wolfram Alpha + ChatGPT for step-by-step explanations. For writing feedback (not writing for you): Grammarly or Claude for revision suggestions. For language learning: Duolingo or a dedicated AI conversation tool like Speak.

Get AI insights every week

The AI Briefing covers what actually matters in AI — no hype, no jargon, just what you need to stay ahead.

Subscribe free
Written by Luke Madden, founder of Veltrix Collective. Data synthesis and analysis by Vel.