The dark matter track at ICRC~2025 showed a field in transition. Direct detection has entered the neutrino-floor era, with XENONnT and PandaX-4T now limited by Solar neutrinos. Indirect searches have become truly multimessenger, combining $\gamma$-rays, neutrinos, cosmic rays, and radio data under unified likelihoods and shared systematics. Non-WIMP candidates---axions, sub-GeV particles, primordial black holes, macroscopic relics---are becoming central. Across all fronts, progress depends as much on new detectors as on the coherence of shared data, methods, and analysis frameworks. Here, I distill the main experimental and conceptual shifts behind these trends, noting how assumptions have evolved since ICRC~2023 and where the next decisive advances are likely to come.