Using recycled aggregates generated from demolition waste for concrete production is a promissory option to reduce the environmental footprint of the built environment. However, predicting the hardened performance of recycled aggregate concrete is one of the main barriers to its intensive deployment in the construction sector. Since traditional empirical approaches are less reliable for predicting the performance of new recycled aggregate formulations, artificial intelligence approaches have been widely developed in recent years towards this aim. In this paper, we conducted an extensive literature review on artificial intelligence (AI) methods that predict the mechanical performance of recycled aggregate concretes and perform sensitivity analysis. The primary methodologies and algorithms found in the literature have been thoroughly described, examined, and discussed in this study concerning their applicability, accuracy, and computational requirements. Furthermore, the benefits and drawbacks of various algorithms have been highlighted. AI algorithms have demonstrated success in a variety of prediction applications with high accuracy. Although these algorithms are robust predictive tools for estimating recycled aggregate concrete's mixture composition and mechanical properties, their performance is highly dependent on data structure and hyperparameter selection. This study could help engineers and researchers to make better decisions about using AI algorithms for mechanical properties prediction and/or to optimise formulations for recycled aggregate concrete.