The quest for advanced superionic materials requires understanding their complex atomic dynamics, but detailed studies of the interplay between lattice vibrations and ionic diffusion remain scarce. Here inelastic and quasielastic neutron scattering measurements in the superionic argyrodite Cu7PSe6 are reported, combined with molecular dynamics (MD) based on ab initio and machine-learned potentials, providing critical insights into the atomistic mechanisms underlying fast ion conduction. The results reveal how long-range Cu diffusion is limited by ntercluster hopping, controlled by selective anharmonic phonons of the crystalline framework. Further, the Green–Kubo simulations reproduce the ultralow lattice thermal conductivity and identify contributions from mobile ions, phonons, and their cross-correlations. The mode resolved analysis shows that the thermal conductivity is dominated by low-energy acoustic phonon modes of the overall crystal framework. The analysis of mode-resolved spectral functions further show that vibrational modes with significant Cu contributions are strongly damped, corresponding to the breakdown of associated phonon quasiparticles. These results high-light the importance of strongly anharmonic effects in superionic systems, in which the traditional quasiharmonic phonon picture is insufficient, and pave the way toward combining machine-learning accelerated simulations with neutron scattering experiments to rationalize the complex atomic dynamics underlying ionic and thermal transport.