You have probably heard the pitch by now: hand a child an AI tutor and the learning takes care of itself. New doctoral research out of Auburn University put a version of that promise to the test, embedding an optional AI assistant inside a real online course. The participants were not children. All 177 were active-duty U.S. Air Force personnel in adult leadership training, and the finding was a quiet one: almost no one used the tool, and for the few who did, it moved nothing the researchers could measure. Their own takeaway is the part worth keeping, because it holds just as well at your kitchen table: access to a tool is not the same as teaching.
New Auburn research on an AI tutor is making the rounds, and it is easy to read into it more, or less, than it says. Here are the questions parents are actually asking about what it means at home.
Common questions
Does giving my child an AI tutor help them learn?
What does the Auburn study actually say?
My child studies with apps and still struggles. What should I do?
Will AI replace teachers?
New Auburn research: an optional AI tutor sat in a real course, and access alone moved nothing. The test was never the AI. It’s whether a tool makes a learner think, and frees the adult beside them.
What the research found
The work is a new doctoral dissertation from Auburn University’s College of Education, completed by Lt. Col. Thomas Outlaw and chaired by Dr. Amy Serafini, and described in an Auburn newswire piece on August 20, 2026. It has not yet appeared in a peer-reviewed journal; the dissertation record is under embargo into 2027. So treat it as early research, attributed to the people who did it, rather than a settled study.
Here is what it looked at. Outlaw built an AI-literacy module and an optional chatbot named Athena into an existing online orientation course, then followed 177 consenting participants. Every one of them was an active-duty member of the U.S. Air Force enrolled in professional military education, the leadership training for enlisted service members. No children, no K-12 classroom. Of the 177 adults, only 10 chose to use the optional chatbot at all. When the researchers compared those 10 against the 167 who did not, they found no statistically significant difference in course performance or in how the learners rated the experience. Access to the tool, on its own, did not move the needle.
Outlaw’s own reading of it was direct. “Access alone does not guarantee meaningful engagement,” he said. His committee chair, Dr. Serafini, drew the line the study kept pointing at: “Even in an era of artificial intelligence, relationships still matter.”
Access alone does not guarantee meaningful engagement.
Laura Lurns · Learning Success expert
The frame the science supports
The easy story to tell about a result like this is that the AI failed. That is not the story, and reaching for it misses the more useful point. The mistake sitting underneath the whole setup is the assumption that the tool is the teaching, that putting an AI within reach of a learner is itself an intervention. It is not. A tool is a delivery vehicle, and what it delivers depends on how it is built and how it is used, not on whether it has AI in the name.
Learning science has a fairly clear read on what actually teaches, and it does not start with the medium. Being made to pull an answer out of your own memory tends to help it hold better than reading it a second time, a pattern that has held up across ages and settings (Roediger and Karpicke, 2006). Practice spread across days tends to stick better than the same work crammed into one sitting (Cepeda and colleagues, 2006). And tutoring that engages a learner at each step, catching where the thinking goes wrong, tends to outperform software that only checks the final answer (VanLehn, 2011; Kulik and Fletcher, 2016). Notice what none of those depend on: a screen, a brand, or the word AI. They describe what a learner is made to do. Two honest limits belong right here. That science was built mostly on human tutors and older tutoring software, before today’s AI chatbots existed, so applying it to an AI tutor is a reasonable expectation and not a proven finding. And the Auburn study itself is thin on this exact point, with only 10 of its adult participants choosing to use the tool, which is a signal, not a verdict.
So the real test of any tool, AI or not, is not whether it is clever. It is whether it makes a learner retrieve, reason, and practice over time, and whether it frees the caring adult to do the work no software has replaced. When a tool skips that and simply keeps someone busy, you get the oldest trap in education wearing a new coat: engagement mistaken for learning. And when a child studies hard, uses the app, and still watches the material slide off, the block often sits a layer below the screen, in an underlying processing skill that no study app was built to notice.
Key takeaways
- Access is not teaching: an optional AI tutor in Auburn’s study moved nothing measurable.
- The design is the test, not the AI: what teaches is retrieval, step-by-step practice, and a caring adult.
- It was adults, not kids: all 177 participants were active-duty Air Force in leadership training.
What it means for your child
You do not need to evaluate an algorithm to make a good call on a study app. Keep one test that outlasts whichever tool is popular this month: sit through a single session and ask, at the end, did my child produce the answers or receive them? A tool that leaves a kid able to work the next problem on their own is earning its place. One that leaves them able only to tap for the next tidy summary is not, however impressive it feels in the moment. That question costs nothing and it works on every product that comes after this one.
Then keep the tool honest about its job. An AI that quizzes and prompts is genuinely useful for a child who is able to learn the material and needs the repetitions. It does nothing for the child whose difficulty sits below the words on the screen, and no amount of slicker software finds that layer for you. If that is the pattern you keep hitting, a look across the systems your child’s learning actually runs on will tell you more than another download. How the process works starts with what you are already seeing at home. A screener is a starting point, not a diagnosis. If your child might need formal accommodations such as an IEP or 504 plan, or you suspect a vision, hearing, or medical cause, pursue a professional evaluation too, since that is the only route to those supports.
And hold onto the part the researchers kept circling back to. Their own words were that relationships still matter, even in an era of AI, and that access on its own guarantees nothing. A well-built tool frees your time. It does not free your child from needing you. The strongest version of all of this is a child recalling, spacing the work out, and doing it beside an adult who knows them and notices when something is off. The apps will keep getting slicker. Your child still needs a teacher, and most days the nearest one is you.
Even in an era of artificial intelligence, relationships still matter.
Laura Lurns · Learning Success expert
You do not need to understand an algorithm to make a smart call for your child. You need one habit: notice whether a tool leaves your child more able to do the work alone, and trust that read over any feature list. Your child is capable of the real thing, the effortful, sometimes slow learning that actually holds. The villain here is not AI, and it is not the quiz or the flashcard, which do honest work. It is the older and stickier idea that putting a tool in front of a learner is the same as teaching them, so a busy screen gets counted as progress while the thinking quietly gets done for the child. That is the trade Learning Success was built to refuse. Our All-Access membership opens an assessment that asks about the processing systems your child’s learning runs on, and a roadmap that names what to build first, so the practice you do together actually pays off. It is a starting point for helping your child, not a diagnosis. If your child might need formal accommodations such as an IEP or 504 plan, or you suspect a vision, hearing, or medical cause, pursue a professional evaluation too, since that is the only route to those supports.
See what All Access gives your child
See MTSS for schools
Is your child struggling in school?
Get your free personalized learning roadmap
You describe what you see at home. We turn it into a plan you start this week.
- Answer 5 short questionnaires about what you already notice, 30–45 minutes at your own kitchen table
- Your child sits no test and gets no score: nothing to schedule, nothing for them to dread
- You do the answering, the AI does the writing, and a person reviews it before it reaches you
- Access all 40+ courses instantly: reading, math, focus, processing and more, with new ones added regularly
Why we use AI, plainly: it writes from a knowledge base our team maintains and audits. We work through it line by line and pull anything the evidence stops supporting. The roadmap you get on Tuesday reflects what we corrected on Monday, and a human still reads it before you do.
Your school district must evaluate your child free of charge if you ask in writing, whatever your income and whatever the outcome (US, 34 CFR 300.111 and 300.301(b)). That route takes time and answers a different question than you do. This one starts today, from what you already know.
Your answers stay yours. We do not sell your personal information, and we do not hand identifiable assessment data to outside AI companies to train their models.
A screener is a starting point, not a diagnosis. If your child might need formal accommodations (an IEP or 504 plan), or you suspect a vision, hearing or medical cause, pursue a professional evaluation too. That is the only route to those supports.
References
- Auburn University — Closing the Distance: how AI can support military learners in distance education
- Auburn University ETD — Scaling Personalized Learning in Higher Education (Outlaw dissertation record)
