The increasing complexity of operations in the automotive supply chain has driven the adoption of advanced technologies to optimize production and supply logistics. In this context, Artificial Intelligence and Machine Learning have emerged as essential tools for enhancing predictability and operational efficiency, enabling more precise and data-driven decision-making. This paper aims to evaluate the impacts of intelligent frameworks in production management and supply logistics of different predictive models applied to demand forecasting and inventory optimization, considering their advantages, challenges, and performance. A methodology combining literature review, case study, and computational modeling to achieve this. Machine learning models such as ANN, LSTM, and GRU were compared to the traditional statistical model Auto-Regressive Integrated Moving Average (ARIMA). Performance was assessed based on mean absolute error (MAE), mean squared error (RMSE), and the coefficient of determination (R2). The Wilcoxon test was applied to verify the significant difference between models before and after the implementation of intelligent frameworks. The results indicated that LSTM and GRU models reduced MAE by 20%, decreased lead time by 30%, and increased logistical efficiency by 18%, contributing to lower operational costs. Despite these advancements, challenges such as computational infrastructure, integration with legacy systems, and organizational resistance must be overcome to maximize the benefits of this technology in the automotive sector.