The Generative AI Generation in Higher Education · Evidence dashboard, Nov 2022 to Nov 2026

Sheet 05 of 17 · Drill-down

Demographics & equity

Adoption is not uniform. Choose a dimension to compare groups. Higher use does not mean greater benefit.

Fig. 5.1: Drill down by dimension

↗

Fig. 5.2: Six layers of the AI divide

Each layer builds on the one below. A student can clear the first and still fall behind at the next, so read from the bottom up. The top layer is about benefit, not access. Tools can be shared equally while learning outcomes stay unequal.

I

Material access Can they get online and use the tools? Devices, connectivity, required tools. Tyton 2024.

About 1 in 4 limited in access to basic infrastructure such as reliable internet (Tyton 2024) ~25%

II

Capability access Do they get the stronger models? Free tiers have tighter limits and fewer features. 11.3% paid at Middlebury. Tyton 2024.

Regular student users who pay for AI tools 44%

III

Instructional access Have they been taught to use it well? Verification, privacy and disclosure. EDUCAUSE 2025.

Received institution-provided AI training 20%

IV

Disciplinary access Do they learn AI within their field? Field norms shape what students practise. Sciences, maths and CS students adopted at nearly twice the rate of literature students at Middlebury.

Academic AI adoption, literature vs. sciences (Middlebury) 49–91%

V

Human support Is a person there when AI falls short? A tutor, librarian, instructor or adviser. This is the layer AI is most likely to replace.

No reliable measure yet —

VI

Outcome equity Do they benefit equally? Learning, persistence and employment. Tutoring gains were largest for students with lower prior knowledge (IZA 2025), but no study yet tracks outcome gaps by group.

No reliable measure yet —

↑ Foundation at the bottom Top layer hardest to measure

Reading note: Higher adoption in a group does not establish greater access to advanced models, stronger literacy or better learning outcomes. First-generation status remains a major evidence gap and should not be inferred from race, income or school type.

Race / ethnicity: Frequent GenAI use for schoolwork (high school)

Pew 2025 found a similar pattern: ~6 in 10 Black and Hispanic teens used chatbots for schoolwork vs ~half of White teens. These are differences in reported adoption and confidence. They say nothing about ability, literacy or outcomes. Comparable undergraduate data are lacking. Source: College Board 2025.

Parental education: Frequent GenAI use for schoolwork (high school)

A clear gradient by parental education, used here as a proxy for socioeconomic context. Source: College Board 2025.

Household income: Chatbot help on all or most schoolwork (teens)

Lower-income teens were nearly three times as likely to rely on chatbots for most schoolwork. That may reflect heavier reliance on free tools where paid tools and human help are harder to get. Source: Pew Research Center 2025.

Achievement: Frequent GenAI use by self-reported grades (high school)

High achievers use AI too, so it is not mainly a shortcut for struggling students. At Middlebury, by contrast, below-median-GPA students had higher adoption (87.1% vs 80.3%). Source: College Board 2025.

Gender: Use by gender, three measures

Gaps are larger for intensity than for any-use, and small in the high-school pipeline. Women report more concern about misconduct accusations and inaccurate results (HEPI 2025). Source: Gallup 2026 · Middlebury 2025 · College Board 2025.

Discipline: Academic adoption by field (Middlebury) and perceived value (Dartmouth)

Field of study was the strongest predictor of adoption at Middlebury. Differences likely reflect task suitability and disciplinary norms, not technical ability. That argues against identical AI-learning goals across majors. Source: Middlebury 2025 · Dartmouth 2026.

Institution type: Two-year vs four-year students

Gallup 2026 found similar weekly use among associate and bachelor’s students; the clearer two-year gap is in guidance, not adoption. Source: Inside Higher Ed 2025.

School access: AI tools available on school network (teacher-reported)

Private schools were also more than twice as likely to permit AI use and have policies. Rural and town schools were less likely to provide network access. Source: College Board 2024–25.