Artificial intelligence has been "about to transform education" since the 1960s. Intelligent tutoring systems produced real but narrow gains over fifty years. Then, from November 2022, large language models (LLMs) put a capable, conversational, free tutor-or-cheating-machine in every student's pocket. This section covers the history of intelligent tutoring, the 2023–26 wave of LLM tutors and teacher copilots, AI grading and feedback, cheating and the limits of detection, data and privacy issues, what the evidence shows so far, and the company landscape. The supporting files work through AI-tutor inference costs and AI-detector base rates, track the main studies, and give a diligence checklist.
What the study covers
- 1Overview
- 2A short history of intelligent tutoring systems
- 3The generative AI wave, 2022–2026
- 4Mechanics: how AI education products work
- 5AI grading and feedback
- 6Cheating and detection
- 7Evidence so far
- 8Data, privacy and policy
- 9Key numbers
- 10Company landscape
- 11Pitfalls
- 12How an expert thinks about it
- 13Glossary
Technical exhibits
Worked examples, checklists and case studies included with this study.
AI Tutor Cost and Detection Math
Inference cost per student under three model-price tiers, with caching; AI-detector base rates and false-accusation arithmetic (computed in Python)
AI Education Evidence Tracker
The main studies of AI tutors, copilots and teacher tools, 2011–2026, with design, sample, effect and caveats
AI Edtech Diligence Checklist
Questions for investors, district buyers and product teams evaluating an AI education product