In this study, Cu–Sn-based diamond composites were fabricated using L-PBF. The effects of the Cu-coated diamond on the melt-pool evolution, diamond–metal matrix bonding, thermal damage, and friction behavior of Cu in the Cu–Sn-based diamond composites were investigated. Moreover, the optimum L-PBF parameters for processing Cu-coated diamond/Cu–Sn composites were investigated. A convolutional neural network based on deep learning was used to process laser-scanned diamond images and evaluate the diamond thermal damage by model-feature visualization. The thermally damaged Cu-coated diamond and percentage of severely damaged parts were reduced compared to those of the uncoated diamond. Thick Cu coating changed the wettability of the diamond and matrix, thereby increasing the friction of the molten pool and unmelted metal particles and introducing the friction of the Cu microfluidic flow on the diamond. Hence, the mechanical holding force of the CuSn alloy matrix on the diamond substantially increased, and the bending strength of the Cu-coated diamond/Cu–Sn composites fabricated using L-PBF reached 720 MPa, which is approximately twice that of diamond/Cu–Sn (365 MPa).