Why does AI tutoring still leave most students behind?
You have seen the headline: AI tutors double learning gains. Then you hear AI cannot replace a human tutor. Both claims come from real research, and both are true.
The Harvard result is real, and narrow
A Harvard and MIT randomized controlled trial found students using an AI tutor produced more than double the learning gains of students in a best-practice active learning classroom, per PYMNTS. The Indiatimes report on the same experiment adds that students reported feeling more engaged and motivated. But this is a controlled trial with motivated Harvard students. It shows the ceiling of what AI tutoring can do, not the average result most schools should expect.
The study pointing the other way
A new study covered by WTOP News finds AI can help kids learn but cannot replace the tutor by itself. Chris Agnew, quoted by WTOP News, says years of research show a human tutor provides engagement and motivation, which leads to better academic outcomes. He calls this high-impact tutoring. The risk he names is cost: with school funding shrinking, districts may be tempted to swap human tutors for AI to save money.
The comparison group is the story
Harvard's trial shows AI alone beating a strong classroom. The study WTOP News reports on shows AI alone falling short of a strong human tutor. Elite students in a trial are not the same population as the learners organizations most need to reach. If you cut human tutors where high-impact tutoring already exists, you likely lose ground. PYMNTS reports Brookings found that when educators co-design AI tools from the start, results shift from passive answer delivery toward teaching that builds skills. In plain terms: the AI stops handing students answers and starts guiding them. The tool itself is not the variable. Who shapes it, and whether humans stay in the loop, is.
What this does not solve
None of these sources measure long-term retention, or outcomes for struggling students outside elite settings. The study reported by WTOP News frames AI as a supplement to human tutors, not a substitute. Cost pressure will push districts toward substitution anyway. No source here tells you how to fund the hybrid model.
Where your organization puts AI in front of learners alone, which people would lose the engagement that only humans provide?
Who on your team would co-design an AI tutor with educators, rather than buying it off the shelf?
List every place your organization uses AI with learners alone, and mark where a human is still in the loop.
Sources
Indiatimes on the Harvard AI tutor experiment
Recommended for you
- Education
Why does plagiarism software still miss actual cheating?
Detection tools compare text against other text. That design choice means they catch copying, not the dishonesty you actually care about.
- Education
What breaks when AI tutors outperform your teachers?
A Harvard-MIT trial shows AI tutoring beats best-practice classrooms. The risk is not the tool, it is who designs how it gets used.
- Education
What breaks when universities grade AI-era essays?
AI writes passing essays, so written assessment no longer proves learning. Universities are falling back on oral exams, but scaling them is the real problem.
- Education
Why does AI widen learning gaps before closing them?
Students use AI everywhere, but almost none are taught how. The gap that matters is guidance, not access, and it splits learners fast.
Enjoyed this article?
Subscribe for new articles. No spam. Unsubscribe anytime.