<p>India, the world’s third-largest CO₂ emitter, faces the critical challenge of balancing rapid economic expansion with its climate commitments under the Paris Agreement. This study develops a hybrid Principal Component Analysis–Artificial Neural Network (PCA–ANN) framework to forecast CO₂ emissions and examine their multidimensional, nonlinear determinants. Annual time-series data spanning 1981–2019, sourced from the World Bank, BP Statistical Review, and national databases, are employed. The model incorporates GDP per capita, fossil fuel and renewable energy shares, urbanisation, international tourism, agricultural output, and a PCA-derived Financial Development Index to address multicollinearity and capture structural financial dynamics. The ANN demonstrates high predictive accuracy (RMSE = 0.018; MAPE ≈ 3.6%; directional accuracy ≈ 98%), confirming robust generalisation. Results indicate that GDP growth, urbanisation, tourism, and agricultural expansion exert significant upward pressure on CO₂ emissions, while renewable energy adoption and financial development exhibit mitigating effects through efficiency gains and green investment channels. These findings reflect nonlinear associative relationships consistent with the predictive nature of ANN models. The study advances the literature by integrating a PCA-based Financial Development Index within a nonlinear ANN framework, modelling a broader set of structural drivers beyond conventional specifications, and offering a system-oriented forecasting approach with direct relevance to India’s SDG 13 targets and Nationally Determined Contributions.</p>