This study explores the application of machine learning algorithms, specifically Random Forest Regressor (RFR) and Gradient Boosting Regressor (GBR), to predict key outcomes of the friction drilling of A356 aluminum alloy. Optimizing process parameters such as rotational speed (RS), feed rate (FR), and preheat temperature (PH) is critical to achieve high-quality bushings during friction drilling. The study focused on predicting bush height (ha), thickness (t), surface roughness (Ra), and the induced temperature at workpiece/drilling-tool interface (T) through a dataset consisting of 27 experiments. The results showed that RS and PH had a significant influence on ha and T, with higher values of both parameters leading to increased bush height and induced temperature. Nevertheless, FR demonstrated a weaker effect on these responses but had a more pronounced impact on t and Ra. Feature importance analysis revealed that RS and PH were the most critical parameters for optimizing the friction drilling process, while FR had a lower effect. Additionally, the GBR model outperformed the RFR model in predicting ha, t, and Ra, providing more accurate results for these dimensions. Whereas the RFR exhibited a better behavior in predicting T, demonstrating the machine learning potential to enhance precision of the formed bushings.