The increasing role of data-driven methodologies in decision-making has led to the integration of Machine Learning (ML) techniques for analyzing Environmental, Social, and Governance (ESG) indicators. This study employs clustering algorithms and time-series forecasting models to assess ESG trends and predict future sustainability outcomes. A comprehensive dataset spanning 2000 to 2020 is utilized to uncover latent patterns, segment ESG indicators based on temporal behavior, and generate forecasts using machine learning-based predictive models. The research integrates data preprocessing, k-means clustering for unsupervised segmentation, and Autoregressive Integrated Moving Average (ARIMA) models for time-series prediction. Results demonstrate meaningful improvements in environmental factors, such as declining CO2 emissions and reduced fossil fuel consumption, alongside progress in social indicators like education. However, persistent challenges in public health and labor market stability necessitate data-driven interventions. The clustering analysis effectively categorizes ESG indicators into stable, moderately variable, and highly fluctuating groups, aiding strategic decision-making. Forecasting outcomes emphasize continued environmental gains but highlight the importance of proactive policy adaptations in social and governance domains. This study underscores the efficacy of Machine Learning in ESG analytics, demonstrating its potential for intelligent sustainability monitoring and policy planning.