<p>In civil engineering constructions, retaining walls (R-walls) are constructed to retain backfill soils, rock, or other materials to maintain terrain stability and prevent erosion or collapse. This study presents reliability analysis (RA) of a gravity R-wall in seismic and static conditions using an efficient hybrid paradigm of extreme learning machine (ELM) optimized with improved grey wolf optimizer (IGWO), i.e., ELM-IGWO. RA was performed for three stability factors including sliding, overturning, and bearing failures. Initially, Monte Carlo Simulation (MCS)-based sampling approach was implemented to generate random samples followed by applying the ELM-IGWO framework to automate the process of RA at different coefficients of variation (COV) levels. The outcomes were compared with six additional hybrid ELMs and four ensemble learning paradigms. As per the results, the proposed ELM-IGWO framework provides the best-fitted estimation for all the stability factors with a correlation coefficient between 0.9902 and 0.9999. These results suggest that the developed ELM-IGWO model demonstrates a consistently high level of accuracy. Overall, the suggested ELM-IGWO framework can be regarded as a viable alternative for assessing the risk of gravity retaining walls in both static and seismic conditions, as well as across COV levels. Thus, a graphical user interface (<Emphasis Type="Underline">GUI</Emphasis>) was developed and attached as supplementary material for future research.</p>