# AI bifurcates and thins the US university system rather than uniformly growing or shrinking it

**Core answer:** Both. AI is an accelerant and sorter of forces already in motion, not the primary cause. Total enrollment drifts from ~19M toward ~14–15M by 2045 (Unverified), but institution count falls faster than degrees conferred — the system thins at the median while the top decile grows. A "barbell" structure emerges: elite and flagship institutions gain share; tuition-dependent, non-selective, teaching-only institutions contract sharply.

**Two-front squeeze:** Fewer 18-year-olds compound with a weakened degree signal. Births fell from 4.32M (2007) to ~3.59M (2023), ~17% decline (CDC/NCHS); high-school graduates peak ~3.9M/year around 2025 (WICHE, *Knocking at the College Door*), and the birth dearth hits college-going age 2029–2037. Simultaneously, AI-assisted skills verification erodes the degree's screening function first; the trust/insurance function (blame-shifting, liability) persists longer.

**Regional asymmetry is real but proxies function, not conference affiliation.** SEC/Sun Belt flagships and land-grants grow 30–50% headcount by ~2045 (Unverified on magnitude), riding expanding cohorts, in-migration, and state funding. Nonselective Northeast/Midwest institutions contract roughly one-third; closures concentrate in PA, OH, IL, New England. But the real dividing line is endowment depth (privates) and state-subsidy commitment (publics