An international research collaboration between the University of Vienna and Lawrence Berkeley National Laboratory has applied machine learning to re-examine the Galactic Center Excess, a diffuse gamma-ray glow surrounding the heart of the Milky Way. Published in Physical Review Letters, the analysis confirms that dark matter cannot be ruled out as the source, significantly weakening arguments that previously favored a stellar origin. The signal has puzzled physicists for more than ten years, with competing theories pointing either to self-annihilating dark matter particles or to a dense concentration of millisecond pulsars. Earlier statistical models favored the pulsar hypothesis, assuming that numerous faint, unresolved point sources could collectively produce the observed emission. The critical limitation of previous studies was the exclusion of photon energy data. The new framework overcomes this by training a machine learning algorithm on more than one million simulated gamma-ray observations, enabling simultaneous evaluation of spatial distribution and spectral energy profiles. Navigating the highly congested and luminous environment of the galactic core required this advanced computational approach to isolate subtle signal patterns. Lead author Florian List of the University of Vienna noted that incorporating photon energy fundamentally shifts the analytical outcome. The revised models demonstrate that if the emission stems from discrete astrophysical objects, those objects must be exceptionally dim. Meeting the observed signal intensity under the pulsar model would require approximately thirty-five thousand such sources concentrated in the galactic center. This estimate vastly exceeds the several hundred to few thousand pulsars currently predicted by standard astrophysical models. Co-author Nick Rodd of Lawrence Berkeley National Laboratory emphasized that these hypothetical sources would be nearly indistinguishable from the diffuse emission expected from dark matter annihilation. The study does not provide direct confirmation of dark matter, but it successfully dismantles one of the strongest statistical objections to the theory. The findings highlight the expanding role of artificial intelligence in high-energy astrophysics, where traditional statistical methods often struggle with complex, overlapping datasets. By integrating previously unused energy metrics into machine learning pipelines, researchers have refined the constraints on competing hypotheses. The revised analysis will inform future observational strategies, guiding next-generation gamma-ray telescopes and theoretical frameworks. As the debate over the galactic center continues, the dark matter hypothesis remains scientifically viable, keeping the search for particle-based dark matter at the forefront of contemporary astrophysical research.