Access is Not Enough: Human Support Improves Engagement with AI Tutoring

Bibliographic Data

Authors
Carly D. Robinson
David Gormley
Ana Trindade Ribeiro
Susanna Loeb
Author(s)
Robinson, C.D., Gormley, D., Ribeiro, A.T., & Loeb, S.
Year of Study
2026
Publication
EdWorkingPapers.com
https://doi.org/10.26300/pz7p-p388

AI tutoring platforms offer a promising path to scaling personalized instruction, but only if students use them. We report findings from two randomized controlled trials in which elementary students were assigned to use an AI literacy platform independently or with an in-person tutor focused on engagement, not direct instruction. Despite dedicated session time, nearly half of students in the control group never used the platform, and those who did averaged only 2-5 minutes per week. Working with human tutors increased average weekly platform usage by 1 to 4 minutes and engagement by 71-80%. However, usage remained low, and the intervention did not improve reading achievement. Findings suggest that access alone is insufficient; implementation and human support matter for promoting meaningful engagement with AI-based learning tools.

Download paper

Research Design

Study Design
Quantitative
Methodology
Randomized Controlled Trial
Subject
Reading
Grade Level(s)
1st Grade,
2nd Grade,
3rd Grade,
4th Grade,
5th Grade
Sample size
355

Program Details

Program Evaluated

AI tutoring with human support

Duration
2024-2025
Student-Tutor Ratio
2-5:1
Topics
AI Tutoring,
Early Literacy

  Media Mentions

| The 74

Artificial intelligence-based tutoring programs promise something schools have long struggled to provide: personalized instruction for every student at a fraction of the cost of traditional 1-on-1 tutoring. 

But for that promise to be fulfilled, students have to use the AI tools. New research suggests that may be much harder to achieve than schools assume. 

In a paper describing the results of randomized controlled trials led by the Stanford SCALE Initiative, 355 students in grades 1 to 5 in two school districts were assigned either to use an AI literacy tutor independently or to use the same platform with support from a human tutor. In both, the goal was to improve reading achievement.

The human tutors encouraged students to participate, helped troubleshoot technology problems and established group norms, while building rapport and accountability. Their sessions followed a consistent structure: a brief check-in, time for students to work on the platform and final reflection. In the group with no human support, students were expected to work on the AI platform on their own. Each group was scheduled to use the tool twice a week for 30 minutes per session.  

The clearest finding from the study was that all students, regardless of whether they received human support, used the AI tutor far less than intended. For example, almost half the children in the independent-use group did not use the AI tutor at all. Those who did use the platform spent just two to five minutes per week, on average.

Students using the AI tutor with human support spent only slightly more time using the tool, but the effects were small: an increase of just one minute a week in one of the districts and 4.4 more in the other. 

There are a range of potential reasons for this low engagement, from students not finding the tool compelling to preferring to ask a teacher for help or needing more encouragement and accountability.

...

| K-12 DIVE

In two districts analyzed by Stanford University, students’ average weekly use of one such tutor was 2.18 minutes and 5.23 minutes, respectively.

Dive Brief:
  • As multiple states prepare or have already implemented artificial intelligence tutoring pilot programs, a new Stanford University study found that merely offering AI tutoring tools to students is not enough to create meaningful engagement and improve reading outcomes.
  • Among two student groups studied, only about 61% in one school district and 53% in another used the AI tutor even when there was scheduled time to use the tool, according to the research released Wednesday. Overall, students’ average weekly usage amounted to 2.18 minutes and 5.23 minutes for each district, respectively.
  • When students used the AI tutoring platform alongside a human tutor, student engagement slightly increased by one minute per week in one district and by 4.4 minutes per week in the other. Still, the researchers said that difference does little to fulfill the 30 minutes of weekly use recommended by the platform provider for measurable reading gains.

...

| The 74

Ed tech companies routinely pitch AI tutoring platforms as a way to deliver personalized instruction at a scale that no human teacher can match. But when researchers from Stanford University looked at how much students actually used one major AI platform, something startling happened: Students didn’t use it that much at all. 

In the study, published Wednesday, two unnamed school districts carved out dedicated time for hundreds of elementary school students to work with a well-known AI reading tutor, either during class time or after school. Researchers followed about 350 students across two randomized controlled trials. All of the students were expected to log on for at least two 30-minute sessions a week.

They found that of the students assigned to work independently with the AI, just over 60% in the first district and 53% in the second ever logged on to the platform — at all.

Among all students, average weekly usage came to just over two minutes in District A and just over five minutes in District B.

Those who did log on averaged 13.2 minutes a week in District A and 25.8 minutes in District B, using the tutoring for just four to five weeks on average in an “intervention window” that ran from 14 to 31 weeks.

...

| Chalkbeat

A group of Stanford University researchers started with one question: Could a human tutor providing motivation and support get students to spend more time working with an AI literacy tutor?

The answer turned out be yes — but only between one and four minutes more per week. Many students never logged on at all.

That left the researchers with a different set of questions.

“A key finding that we weren’t even meaning to test is that having access to this AI tutor isn’t the same as using it,” said Carly Robinson, the lead author on the study released Wednesday and the director of research for the SCALE Initiative at the Stanford Accelerator for Learning.

Robinson said this doesn’t necessarily mean AI tutoring doesn’t work. “We never really got close enough to the dosage needed to find out.”

The study adds to a very limited research base for AI tools in education as many school districts are looking for ways to maintain or increase tutoring that costs less than hiring people to do the work. It also aligns with the experience of Khanmigo founder Sal Khan, who recently acknowledged that students didn’t engage much with the AI tutor he initially hoped would revolutionize education.

Robinson and her colleagues worked with two school districts serving high-poverty populations using the same AI learning platform in different tutoring settings. Neither the districts nor the learning platform are named in the study.

In one school district, elementary children were supposed to use the literacy tutor during homework time in an afterschool program. In the other, early elementary students were supposed to use the AI tutor during class time.

The learning platform provider said that students would show improvement in their reading skills with at least 30 minutes of use a week. The goal was for students to complete two 30-minute sessions each week.