The case against ΛCDM rests on three pillars of modern observational cosmology: the clustering of galaxies, type Ia supernovae, and the Cosmic Microwave Background. Each probe measures a different aspect of the universe’s expansion history and growth of structure, and each comes with its own systematics. When combined, they constrain the equation-of-state parameter w, the ratio of dark energy’s pressure to its energy density. A pure cosmological constant corresponds to exactly w = −1, unchanging for all of cosmic time. Any statistically significant deviation from that value, or any evidence that w evolves, would amount to a conviction that dark energy is dynamical rather than constant.

The most prominent evidence has come from the Dark Energy Spectroscopic Instrument, or DESI, which measures baryon acoustic oscillations, the fossil imprint of sound waves in the early universe that serves as a standard ruler for mapping expansion. DESI’s first data release, published in 2024, and its second release in 2025, when combined with supernova and microwave background data, have shown hints that dark energy may be weakening over time. In the widely used Chevallier–Polarski–Linder parametrization, which allows w to vary as w(a) = w0 + wa(1 − a), the combined data prefer values that drift away from the ΛCDM point of w0 = −1 and wa = 0. The statistical significance of these hints depends on which supernova sample is used, but in some combinations it reaches the three-sigma level, enough to make the community sit up and take notice.

Supernova data themselves are under scrutiny. The Dark Energy Survey’s full five-year sample of roughly 1,500 high-redshift type Ia supernovae, together with the Pantheon+ and Union3.1 compilations of around 2,000 supernovae, have each been analyzed for evidence of evolving dark energy. A reanalysis of the DES supernova program with an updated calibration reported evidence for evolving dark energy, while other work has emphasized possible systematics. One particularly consequential issue is progenitor age bias: analyses have found strong evidence that the properties of type Ia supernovae correlate with the ages of their host galaxies, an effect that could mimic or mask cosmological signals. Counter-analyses, however, argue that supernova cosmology remains robust to host galaxy age evolution and that the universe is still accelerating regardless of how the dark energy debate is resolved. As Lahav and Shah note, the jury on supernova systematics is still deliberating.

The Cosmic Microwave Background adds its own complications. The Planck 2018 results remain strikingly consistent with ΛCDM, and newer measurements from the Atacama Cosmology Telescope DR6 and SPT-3G broadly confirm the picture, though with intriguing wrinkles. One wrinkle involves the optical depth to reionization, a parameter that governs how the first stars reionized the intergalactic medium. Some analyses suggest that an increased optical depth could relieve tensions within the CMB data, and related work has explored how inferences about neutrino masses, which have in some fits come out unphysically negative, might be entangled with the optical depth. Meanwhile, the flood of galaxies discovered by the James Webb Space Telescope has raised questions about whether there are enough early photons to account for reionization on the usual timescales, prompting proposals of rapid late-time reionization scenarios. Each of these threads shows how a single parameter adjustment can ripple through the entire inferential chain.

Weak gravitational lensing surveys, which measure how the clustering of matter distorts the images of distant galaxies, provide an independent test of ΛCDM’s predictions for the growth of cosmic structure. Results from the Kilo-Degree Survey, the Dark Energy Survey, and the Hyper Suprime-Cam have historically shown mild tension with Planck’s predictions for the amplitude of matter clustering, a discrepancy known as the S8 tension. The latest KiDS-Legacy analysis, using the complete survey, has moved the lensing constraints closer to the Planck values, weakening what once looked like a serious crack in the model. The DES Year 6 clustering and lensing results are expected to sharpen this picture further. For the prosecution, the fading of the S8 tension is an inconvenient development; for the defense, it is evidence that earlier anomalies were statistical fluctuations or systematic errors rather than signs of new physics.

Beyond the raw data, the debate has a statistical dimension that Lahav and Shah treat with care. Bayesian model comparison asks not merely which model fits the data best, but whether the improvement in fit justifies the additional parameters that dynamical dark energy models require. Several recent analyses have examined the DESI evidence for evolving dark energy from a Bayesian perspective, quantifying tensions and comparing standard cosmologies, and the conclusions range from cautious skepticism to modest support for evolution. There is also a deeper theoretical problem: as work on the underdetermination of dark energy has shown, many physically distinct models, including scalar fields with nonminimal couplings, theories that cross the so-called phantom divide where w < −1, and models that violate the null energy condition, can all be made consistent with current data. The mirage of a measured w exactly equal to −1, as Eric Linder warned nearly two decades ago, means that even a perfect agreement with ΛCDM would not prove the cosmological constant is real.

The trial, then, is far from over, and the next decade promises a flood of new evidence. Stage-5 spectroscopic surveys, including the continued operation of DESI and proposed facilities such as the MUltiplexed Survey Telescope and the Wide-field Spectroscopic Telescope, will map tens of millions of galaxies with unprecedented precision. The Nancy Grace Roman Space Telescope, the Vera C. Rubin Observatory’s Legacy Survey of Space and Time, and dedicated supernova programs such as TiDES on 4MOST and the Dark Energy Bedrock All-sky Supernova Program will multiply the supernova sample and control its systematics. The Simons Observatory, SPT-3G, and the LiteBIRD satellite will refine the CMB constraints, while the NASA Landolt Mission aims to deliver a new generation of fundamental photometric standards that could recalibrate the entire supernova distance ladder. Peculiar velocity surveys, such as the DESI Fundamental Plane program, offer yet another route to measuring cosmic expansion that is less dependent on supernova standardization.

What is at stake goes beyond a single parameter. If dark energy is truly constant, the universe will continue to expand forever at an accelerating rate, drifting toward a cold, empty future in which galaxies recede beyond each other’s horizons. If dark energy weakens, decays, or eventually reverses sign, the long-term fate of the cosmos could be very different, and the theoretical machinery of general relativity may need modification on the largest scales. Either outcome would be profound. A confirmation of Λ would deepen what physicists call the cosmological constant problem, the baffling discrepancy between the observed value of Λ and the enormous predictions of quantum field theory, arguably the worst quantitative prediction in the history of science. A detection of dynamical dark energy would point instead toward new fields or modified gravity, reshaping particle physics and cosmology alike.

Lahav and Shah’s verdict is deliberately measured. The recent measurements from galaxy clustering, supernovae, and the CMB have indeed put ΛCDM on trial, and some combinations of data show suggestive evidence against a constant Λ, but the case is not yet proven. Systematic uncertainties in supernova calibration, host galaxy evolution, reionization physics, and CMB foregrounds all remain active areas of investigation, and the history of cosmology offers many examples of anomalies that dissolved under scrutiny. At the same time, the convergence of independent hints from DESI and multiple supernova samples is unusual enough that the community is right to prepare for a paradigm shift. The coming years of Stage-5 surveys and space-based observations will act as the appeals court, and whichever way the evidence falls, our understanding of the universe’s most mysterious component, and of the universe’s own destiny, is set to be transformed.

Subject of Research: Testing the ΛCDM cosmological model and the evidence for dynamical dark energy from galaxy clustering, supernovae, and the Cosmic Microwave Background

Article Title: ΛCDM in a courtroom

Article References: Lahav, O., & Shah, P. (2026). ΛCDM in a courtroom. Nature Astronomy. https://doi.org/10.1038/s41550-026-02977-5

Image Credits: AI Generated

DOI: 10.1038/s41550-026-02977-5

Keywords: cosmology, dark energy, cosmological constant, ΛCDM, DESI, type Ia supernovae, Cosmic Microwave Background, baryon acoustic oscillations, weak gravitational lensing, Hubble tension, Bayesian model comparison, Stage-5 surveys