Abstract:
This study provides a comprehensive intellectual landscape of the integration of
artificial intelligence (AI) with carbon accounting and emissions management.
Despite the growing integration of AI in sustainability practices, a significant
knowledge gap exists in understanding the systematic evolution and methodological
progression of AI applications in carbon accounting. The objective of this study is to
map the research trajectories and evolving paradigms of how AI technologies support
decarbonization, emissions monitoring, and sustainability reporting, and to provide
evidence-based guidance for future studies. A systematic bibliometric analysis was
conducted in accordance with PRISMA guidelines, using the Web of Science and
Scopus databases. The study covers 1,062 research publications spanning three
decades from 1994 to 2025. It employed VOS viewer, Biblioshiny, and R-based
bibliometric tools. Analyses included co-word networks, thematic evolution, and
strategic positioning using density-centrality frameworks to identify the intellectual,
conceptual, and methodological dynamics. The study highlights the AI applications
in carbon accounting and emissions management across three phases: 1994-2020:
Early experimentation with neural networks and footprint modeling; 2021-2023:
Methodological diversification with machine/deep learning and life-cycle
assessment, focusing on themes of neutrality and decarbonization; 2024-2025:
Convergence on ESG reporting, emission monitoring, and net-zero strategies. The
literature is heavily concentrated in leading research economies such as China, the
USA, and India. Findings highlight the need for cross-disciplinary integration of AI
with carbon governance, ESG disclosure, and sustainability strategies. Future studies
should explore the ethical dimensions of AI-enabled carbon accounting, including
fairness, transparency, and credibility in emissions measurement, disclosure, and
governance.
