The most useful AI learning tools do something a busy classroom rarely can: they respond to each student, one step at a time. A teacher with 28 students cannot sit beside every one of them through every problem, and the best AI-enhanced learning platforms are built to fill some of that gap. The research has arrived quickly, and it carries a warning along with the promise. Tools that make students do the thinking have produced large gains in recent trials, while tools that simply hand over answers have left students worse off once the tool was taken away.
Below are ten kinds of AI tools for personalized learning, with named examples of each, followed by what the recent research shows and what a school should check before adopting one.
10 types of AI tools for enhancing learning
1. AI tutors and chatbots
An AI tutor answers a student’s questions in conversation, explains a step that did not make sense and asks the student to try the next one. The good ones are designed not to give the answer away. Khanmigo, from Khan Academy, is the best-known example built for schools: it guides students with questions and hints, and its teacher tools are free for U.S. teachers. Students under 18 can use it only through a school district partnership, a parent-linked account or a writing task a teacher assigns.
2. Adaptive learning platforms
An adaptive learning platform chooses the next question based on how a student answered the last few. A student who is struggling gets more practice on the skill underneath; a student who has mastered it moves on. Examples include ALEKS from McGraw Hill for math and chemistry, DreamBox Math from Discovery Education, and IXL, which raises or lowers the difficulty of questions as a student works.
3. Intelligent tutoring systems
Intelligent tutoring systems are the older, more structured cousin of the AI chatbot. They follow a student through each step of a problem, recognize the specific mistake a student made and adjust what comes next. Carnegie Learning’s MATHia is the most studied commercial example in math.
4. Language learning with AI feedback
Language tools now use AI to explain mistakes and hold practice conversations. Duolingo Max, launched in 2023, added Explain My Answer, which tells a learner why a response was right or wrong, and Roleplay, a practice conversation followed by feedback. For English learners and world language classes, that kind of feedback fills time a teacher cannot give each student individually.
5. Smart content creation for teachers
Some of the most widely used AI in schools never reaches a student directly. Tools such as MagicSchool and Brisk Teaching generate quizzes, leveled readings and study guides from a teacher’s prompt, so the same lesson can be offered at several reading levels. edCircuit covered both in its list of AI tools for elementary students.
6. AI-assisted grading
AI-assisted grading saves time on the most repetitive part of teaching. Gradescope, owned by Turnitin since 2018, uses AI to sort similar answers into groups so a teacher can grade a whole group at once. The teacher still decides the score, which is the right division of labor: AI handles the sorting, and the professional makes the judgment.
7. Personalized study plans
Study-plan tools build a review sequence from what a student has and has not yet mastered. Khan Academy’s course mastery system and Quizlet’s Learn mode both work this way, bringing back the material a student missed and spacing out what the student already knows.
8. Data analytics for student progress
Analytics tools look across assignments, attendance and assessments to show which students are falling behind and where. Most schools get this through the dashboards in their learning management system or student information system, or through student-success platforms such as Panorama Education. The value depends on what happens next: a dashboard that flags a struggling student helps only if someone has time to act on it.
9. Gamified practice
Game-based tools keep students practicing longer. Kahoot! generates quiz games from a teacher’s own material, and Prodigy wraps math practice in a role-playing game that adjusts to each player. They work best for review and fluency, not for teaching a concept the first time.
10. Virtual reality for immersive learning
Virtual reality lets students explore places and processes they could not reach in a classroom, from the inside of a cell to a historic site. ClassVR from Avantis Education is one example built for schools, packaging headsets with ready-made lessons that a teacher controls from a single screen. VR is the least common tool on this list, and the hardware, charging and supervision it needs are real costs for a school.
What recent research says about AI tutoring
Three randomized trials from 2024 and 2025 show what well-designed AI tutoring can do. At Harvard, students in an introductory physics course learned more than twice as much from an AI tutor as from an active-learning class taught by experienced instructors, and took less time doing it, according to the study published in Scientific Reports in June 2025. The tutor was built to reveal one step at a time and to have students attempt each step before it helped.
A World Bank trial in Edo State, Nigeria, gave first-year senior secondary students six weeks of after-school English tutoring with Microsoft Copilot, built on GPT-4. Scores rose by 0.31 standard deviations overall and 0.23 in English, gains the researchers equate to 1.5 to 2 years of typical schooling. AI can also make a human tutor better. In a Stanford trial in spring 2024, elementary students whose tutors used Tutor CoPilot, an AI assistant that suggests how to respond to a student, were 4 percentage points more likely to master math topics, and up to 9 points more likely when their tutor was among the less experienced. It cost about $20 per tutor per year.
When AI tools can harm learning
The same research shows the risk. In a field experiment with nearly a thousand high school math students, published in PNAS in 2025, students given an ordinary GPT-4 chat interface scored 48 percent higher on practice problems while they had it, but scored 17 percent lower than students who never had access once the tool was taken away for the exam. Students given a version designed to act as a tutor, with safeguards against handing over answers, avoided that loss. The authors concluded that students had used the unrestricted chatbot as a crutch. The common thread across all four studies is design: tools that make students do the thinking help them learn, and tools that do the thinking for them do not.
What schools should check before adopting an AI learning tool
The Children’s Online Privacy Protection Act (COPPA) requires parental consent before an online service collects personal information from a child under 13, and schools may consent for parents only for educational use under the school’s direction. Updated federal rules that took full effect in April 2026 also require companies to get separate consent before sharing a child’s data with third parties and to keep a written policy on how long they retain it, according to the amended rule.
Before adopting a tool, a school should confirm a signed data privacy agreement with the company, decide whether students or only teachers will use it, and know what the company does with what students type. It should also ask the question the research raises: does this tool make students do the work, or does it do the work for them? Teachers are mostly answering these questions on their own. Only 18 percent of teachers had received formal guidance from administrators on how to use AI tools, a 2026 Gallup survey found, and a 2025 Gallup and Walton Family Foundation survey found that teachers in schools with an AI policy saved more time with AI than teachers elsewhere.
AI tools help most when teachers direct them
AI learning tools are no longer an experiment for early adopters. The strongest of them can give a student patient, step-by-step help that a teacher with a full class cannot always provide, and the evidence for well-designed tutors is now stronger than for most classroom technology. The weakest can quietly do students’ work for them. The difference rarely shows in a product demonstration, so it falls to teachers and school leaders to choose tools that ask more of students, set clear rules for their use, and keep watching whether students are learning or only finishing.
Related reading
Subscribe to edCircuit to stay up to date on all of our shows, podcasts, news, and thought leadership articles.
More on Artificial Intelligence
