Another Fermi-LAT Dead End

The Fermi-LAT collaboration looks for signals of dark matter annihilations. All of its previous potential signals have turned out to have other more plausible explanations than dark matter annihilation. The latest blip in its data set is no different.

A narrow gamma-ray line would be a distinctive signature of dark matter, but its interpretation depends as much on its spatial distribution as on its energy. 

We reanalyze the reported 43 GeV feature in Fermi-LAT observations of Virgo, Fornax, and Ophiuchus. The excess is reproduced in the same public data and survives event-type partitions, a test using data withheld from the original selection, and detector-coordinate permutation tests. Realistic hadronic and inverse-Compton emission is too broad to explain it. A spatial--spectral likelihood analysis restricted to Good Time Intervals and including diffuse backgrounds and catalogued sources finds a broad line-like component. Uniform surface brightness and substructure-enhanced annihilation are preferred, whereas smooth NFW annihilation and a central point source are unlikely explanations. An independent annular analysis agrees: the events associated with the excess extend through much of the virial region. Calibration over 20--70 GeV and five morphologies gives a conditional global significance of approximately 2.7σ, peaking at 44.5 GeV for uniform brightness. This result is conditional on the preselected three-cluster sample. 

The broad morphology is the main obstacle to a dark-matter interpretation. Such an interpretation would require photons to be a leading annihilation channel, annihilation to remain efficient in cold subhalos, and subhalos to supply nearly all the cluster luminosity. Yet a boost consistent with ΛCDM drives the required two-photon cross section into the range most tightly constrained by Galactic-halo line limits. The most likely explanation is a chance fluctuation amplified by analysis and target-selection effects, possibly compounded by residual Galactic diffuse or unresolved-source mismodeling; if celestial, its morphology makes a simple dark-matter interpretation unlikely.

Stefano Profumo, "Right Energy, Wrong Profile: Why the 43 GeV Cluster Line Is Unlikely to Be Dark Matter" arXiv:2609.16425 (September 14, 2026).