The universe must be older than its contents, and ΛCDM keeps shaving the margin thin. The CMB-calibrated age is 13.79 +/- 0.06 Gyr, while the oldest globular clusters are dated at 12.5 to 13.5 Gyr and the metal-poor subgiant HD 140283, the Methuselah star, carried early estimates of 14.5 +/- 0.8 Gyr before refined modeling brought it to 12.0 +/- 0.5 Gyr (arXiv:2105.11311, arXiv:2411.12343). Several recent analyses still infer cluster and halo-star ages crowding or nominally exceeding 13.8 Gyr within their errors (arXiv:2401.11549). The margin between the oldest objects and the cosmic age is at best a few hundred million years, into which all of star formation's first epoch must fit.
The Hubble tension turns this squeeze into a vice. The 13.79 Gyr age assumes the Planck expansion history with H0 = 67.4; if the local measurement of 73 km/s/Mpc is the true global rate, the universe in flat ΛCDM is only about 12.6 Gyr old, younger than the best-fit ages of the oldest globular clusters and of Methuselah itself. The age crisis of the 1990s, supposedly resolved by dark energy, returns through the front door: the model cannot simultaneously accept the local expansion rate and the stellar chronometers. Every proposed late-time solution to the Hubble tension inherits this problem, since raising H0 globally lowers the age below what stellar physics allows.
The standing is sharpening as both sides improve: Gaia parallaxes and interferometric radii are shrinking stellar-age errors toward the few-percent level, and JWST is pushing globular-cluster formation earlier. Stellar ages are the one cosmology-independent clock in the sky, and they currently sit closer to the Planck-age answer than the local-H0 answer, an arbitration the model itself cannot perform.