Classroom AI Snapshot 2026

What subjects are adopting AI for teaching?

As the school year winds down, we looked back at the last three months of QuestionWell usage data to see where classroom AI is actually showing up. To create this report, we automatically annotated each activity with a subject and analyzed the counts. Test prep includes content that is clearly about standardized test practice or exam prep, since it falls outside of ordinary "classroom" use.

Subjects

We also grouped subjects into four clusters: STEM, humanities, applied/specialist and test prep.  STEM and humanities are nearly tied. STEM subjects represent 41.0% of labeled activity, while humanities and literacy represent 39.9%.

The Long Tail of Topics

But the most interesting story may be the long tail.

The data includes prokaryotic cells, photosynthesis, convection, the Mexican Revolution, Indonesian labor law, Thai geography, Chinese literature, Vietnamese text cohesion, alliteration, Arabic AI lessons, web design, fractions, substance education, medical terminology, supply and demand, leadership skills, Tamil, and Spanish verbs.

That messy specificity feels like real learning.

Teachers are not just asking AI for generic worksheets. They are bringing the actual material of their classrooms: local standards, multilingual content, regional history, niche readings, textbook excerpts, and lessons that do not fit neatly into a stock content library.