Pile-supported structures commonly found in both offshore (e.g. offshore oil and gas platforms) and coastal environments (e.g. sea bridges, piers and jetties) are generally built by means of a group of piles in different arrangements. The correct prediction of the wave loading of closely-spaced piles of these structures is vital for both safety and economical viewpoints. Unlike single isolated piles, where a large number of studies are available together with the well-known Morison equation which is still widely applied for the calculation of wave-induced force, less research studies have been made on wave-pile group interactions. In fact, no reliable wave load formula is yet available for the prediction of wave-induced forces on a slender pile, for which the pile diameter (D) is generally less than about 0.2 times the wave length (L), within a pile group. In this study, new wave load formulae for the prediction of wave-induced force on a slender pile in pile groups with different arrangements are developed using a series of laboratory data obtained from systematic model tests conducted in the 2 m-wide wave flume of Leichtweiss-Institute for Hydraulic Engineering and Water Resources (LWI) in Braunschweig, Germany. For the analysis of the laboratory data and the development of the new prediction formulae, an artificial intelligence (AI)-based computational tool, named "hybrid M5MT-GP model", is implemented. The new hybrid model and the new wave load formulae allow us to systematically assess the pile group effect (K G ) as a function of the flow regime (KC number) and the relative spacing (S G /D) for each tested pile group arrangement. The results show that the pile group effect needs to be considered in calculating wave loads on the slender piles in pile groups, unless K G = 1 where there is no interference effect between neighbouring piles and piles in the group can be treated as a single isolated pile. The accuracy of the new formulae in predicting pile group effect K G is confirmed by the statistical indicators using agreement index I a , correlation coefficient CC and scatter index SI.