With the growing global emphasis on carbon neutrality, carbon capture, utilization, and storage (CCUS) has emerged as a critical pathway in climate mitigation. Meanwhile, advances in artificial intelligence (AI) offer transformative tools to enhance CCUS performance across capture, utilization, storage, and monitoring stages. This study conducts a comprehensive bibliometric analysis to track the evolution of AI-enabled CCUS research from 2001 to 2025. Based on journal distributions, international collaboration networks, and keyword dynamics, four evolutionary stages are identified: (1) auxiliary tool usage, (2) multi-scenario experimentation, (3) integrated modeling, and (4) intelligent system integration. We further delineate the co-evolution of AI algorithms (from SVM/ANN to GNN, PINN, and Transformer) and policy drivers (e.g., IPCC reports, Paris Agreement). Key technical challenges are identified, including data heterogeneity, limited interpretability and physical consistency of black-box models, geoscience-AI semantic mismatch, and deployment regulation gaps. In response, this paper proposes future directions involving open multi-source datasets, hybrid models with physical constraints, ontology-guided semantic embeddings, and scalable containerized deployment frameworks. This work aims to offer a systematic roadmap for AI–CCUS integration, supporting its transition from research demonstrations to scalable industrial deployment.
Ma et al. (Wed,) studied this question.