NSF-funded project explores the use of AI-powered chatbots to prepare future educators

AI-powered tutoring uses machine learning to deliver adaptive, personalized instruction that responds to each student's answers in real time. Unlike static content, AI tutors adjust difficulty, offer hints, explain misconceptions, and track progress across sessions. Leading examples include Khan Academy's Khanmigo, Carnegie Learning MATHia, and Duolingo.

AI-powered tutoring closes learning gaps significantly compared to traditional classroom instruction, but controlled studies show expert human tutors still produce larger gains when sessions are one-on-one. AI tutoring's advantage is availability and scale: it can serve every student simultaneously at any hour, which human tutors cannot.

AI-powered tutoring is most mature in mathematics, reading comprehension, writing feedback, coding, and foreign languages — areas with clear correct answers or established evaluation rubrics. Platforms like IXL, Carnegie Learning, and Duolingo cover these subjects with high accuracy; open-ended humanities discussions remain less well-served.

AI tutoring ranges from free to moderately priced. Khan Academy and Khanmigo are free or low-cost for individuals. Duolingo's premium tier is under $15 per month. Carnegie Learning and similar school-grade platforms require district licensing. AI tutoring is dramatically cheaper than private human tutoring, which averages $40-100 per hour.