The early universe is overlit. JWST's deep surveys have measured the rest-frame ultraviolet luminosity function, the census of star-forming galaxies by brightness, at redshifts 9 through 14 and beyond, and the bright end refuses to decline on schedule: JADES, CEERS, COSMOS-Web, and UNCOVER consistently find bright galaxies at z above 10 in excess over essentially all pre-JWST theoretical models, with the discrepancy growing toward higher redshift and reaching factors of several to ten at z of 12 to 14 (Whitler et al. 2025; Adams et al.; the UNCOVER overabundance analyses). The luminosity function's shape flattens at the bright end, and its characteristic density declines far more slowly with redshift than halo-growth-driven models predict: the same number density of UV-bright galaxies persists across an epoch in which the underlying halo population should be collapsing exponentially.
The model's repair options each carry a signature: elevated star-formation efficiency at early times (why would feedback weaken exactly then?), top-heavy initial mass functions boosting UV light per unit mass (ASTRAEUS and others now invoke this at z above 10), stochastic burstiness scattering faint galaxies brightward (helps at z of 9 to 11, struggles at 13+), hidden AGN contributions, and dust vanishing from bright sightlines. Spectroscopy has confirmed enough of the photometric candidates, including the z of 14 systems with their chemical maturity, to close the contamination escape; the excess is real and the explanations are multiplying faster than they are discriminating.
The standing is one of JWST's defining results: the bright early universe is now measurement, not claim, and the UV luminosity function's slow evolution at z above 10 is a direct, statistical challenge to growth-limited structure formation, awaiting the wide-area verdicts of Roman and Euclid.