At a global level, the analysis of the future today represents a strategic alternative that has been accepted by organizations in order to reduce uncertainty, reorient and strengthen decision-making, reduce risks and configure plausible scenarios; to this end, Strategic Prospective (SP) studies constitute the most widely applied alternative, however, Artificial Intelligence (AI) and its Machine Learning (ML) techniques have also been accepted. SP offers a qualitative approach based on trends, key uncertainties, implications and expert opinion, while ML emerges as a quantitative tool for the exploitation of historical data and identification of hidden patterns dedicated to predicting complex phenomena with high precision, which is why it has been considered of high strategic value. Considering the potential of both approaches, the present systematic review emerged, which aims to identify prospective studies that applied ML to strengthen their results, then offer the reader a current overview that allows them to exploit knowledge through the strategic integration of innovative tools. The review allows us to answer the questions: Has the potential of ML been exploited during the development of SP? In what areas have both approaches been integrated? What are the most preferred ML algorithms that have shown the best results? During the review, 3 expert researchers participated who selected 48 articles published in indexed journals in the last 5 years, applying eligibility criteria, the analysis of 11 were concluded in the areas of: health, education, finance, mining and archaeology, the preferred algorithms were: Artificial Neural Networks (ANN), Random Forest (RF) and logistic regression, however, the best performance was Naibe Bayes with .99, boosted gradient machine with .96, multivariate regression with .95 and ensemble generalized linear model with .94. Limitations of ANNs in financial analysis and RF in health were observed. To validate results, the metrics were applied: confusion matrix, f1 score, area under the curve, confidence interval, Gini coefficient, performance was in the range of .75 to .99. SP and ML have opposite approaches and correspond to different areas of knowledge, which is why the number of authors who address them together is small. However, they are complementary to the task of strengthening SP studies and providing reliability through analysis and quantifiable results.