In this paper, a multi-objective artificial bee colony (MOABC) optimization algorithm is utilized to improve the performance of an adaptive robust control technique. This approach is implemented on an inverted pendulum system. More precisely, the proposed controller is a combination of a decoupled sliding mode controller (DSMC) and adaption laws based on the gradient descent approach. In order to achieve the optimum control operation, the MOABC, as a novel meta-heuristic method simulated from the smart foraging activity of honey bee groups, is employed to optimize the coefficients of the suggested controller. In this regard, the objective functions are determined as the integral time of the absolute value of the pole angle and cart position errors. Finally, the time responses of the system states and control effort are presented to prove the effectiveness and feasibility of the suggested strategy compared to other contemporary studies referenced in the paper.