Abstract: Hydrogen embrittlement poses a critical challenge limiting the widespread application of high-strength steels. Although introducing nanoscale precipitates and elemental segregation at grain boundaries (GBs) has emerged as a promising strategy to enhance material performance, the atomic-scale mechanisms by which alloying elements influence hydrogen behavior through GB segregation and precipitate interactions remain insufficiently understood. To address this gap, this study integrates a custom-developed deep potential (DP) with Monte Carlo (MC) and molecular dynamics (MD) simulations to systematically investigate the synergistic effects of elemental segregation and hydrogen diffusion in Fe–Mn–V–C–B–H multi-component alloys. The developed DP model exhibits exceptional consistency with density functional theory (DFT) in predicting GB energies, GB segregation energies, and elastic constants, significantly surpassing conventional empirical potentials. MD simulations reveal that manganese (Mn) segregation at GBs substantially suppresses hydrogen diffusion, with the inhibitory effect intensifying as Mn concentration increases. Simultaneously, the vanadium carbide/alpha-iron (VC/α-Fe) interface, characterized by periodic lattice matching, functions as an efficient hydrogen trap, effectively localizing hydrogen atoms and reducing their diffusion within the matrix. This work elucidates atomic-scale mechanisms of hydrogen embrittlement in multi-component alloys, highlighting the efficacy of machine-learning potentials (MLPs) for guiding targeted elemental segregation and precipitate engineering to develop hydrogen-resistant structural materials.