Abstract: The reliability of conductive slip rings, which are key components of the solar array drive assembly (SADA), is critical to the satellite’s service life. Traditional analytical models struggle to characterize the complex nonlinear relationship between slip-ring wear and reliability. To address this issue, an artificial intelligence model integrating Tuna Swarm Optimization and Gaussian Process Regression (TSO-GPR) was proposed and trained using ground wear test data. An interdisciplinary equation was established to map the wear rate to the reed pressure and the hardness of the brush block material, and a belief reliability model was subsequently developed to account for uncertainties from multiple parameters. Validation results indicated that the TSO-GPR model reduced the root mean square error (RMSE) and mean absolute error (MAE) by approximately two orders of magnitude compared to the conventional GPR model, demonstrating significantly improved predictive accuracy and generalization capability for conductive slip-ring life prediction. Furthermore, sensitivity analysis revealed that the hardness of the brush block material had a greater influence on reliability than the reed pressure. These findings provide a useful reference for the design of high-reliability conductive slip rings.