Generative Artificial Intelligence is changing production in an accelerating way. However, given that it has a relatively high demand for computing power and energy consumption, it generates a significant amount of greenhouse gas emissions. Whether it will be a “carbon increaser” or a “carbon reducer” in the process of green transition is still an urgent problem to be solved. This paper will study the carbon emission intensity and the green spillover effects of generative artificial intelligence, determine its environmental costs and green benefits, and provide reference materials for policymakers in their green development plans for the digital economy. From a life-cycle perspective, this paper sets the scope of measurement as hardware production, model training and cloud-based inference, and suggests a system for carbon intensity measurement and analytical models. Select representative large-scale models for empirical estimation. At the same time, a theoretical system explores the transmission paths of green spillover effects and carry out corresponding empirical studies. Based on the above data, the carbon intensity of generative artificial intelligence suggests substantial changes in both scale and structure. Most of the carbon emissions occur during the training stage, and although the proportion from the inference stage has been increasing over time, it is still relatively small. The regional energy structure of the models’ deployment appears to be associated with carbon intensity. The results suggest a potential nonlinear pattern between generative artificial intelligence development and regional carbon intensity, although the direction of this relationship remains conditional on energy structure, digital infrastructure, and environmental regulation. The realization of green spillover effects depends on energy structure, digital infrastructure and environmental regulations. Therefore, this paper puts forward some policy recommendations, such as constructing a normalized carbon accounting system, optimizing the spatial arrangement of computing facilities, establishing integrated development mechanisms, enhancing incentive policies, and strengthening international cooperation. The scope of the study is broadened to include the environmental impact of the digital economy, and both theoretical and practical support for the green development of intelligent technologies and low-carbon industrial transformation will be offered.
- Open Access
- Article
Carbon Intensity Measurement and Green Spillover Effects of Generative Artificial Intelligence
Author Information
Received: 23 May 2026 | Revised: 28 Jun 2026 | Accepted: 28 Jul 2026 | Published: 10 Aug 2026
Abstract
Keywords
generative artificial intelligence | carbon intensity | green spillover effects | energy efficiency
References
- 1.
Goldfarb, A.; Tucker, C. Digital Economics. J. Econ. Lit. 2019, 57, 3–43.
- 2.
Strubell, E.; Ganesh, A.; McCallum, A. Energy and Policy Considerations for Deep Learning in NLP. arXiv 2019, arXiv:1906.02243.
- 3.
Henderson, P.; Hu, J.; Romoff, J.; et al. Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning. J. Mach. Learn. Res. 2020, 21, 1–43.
- 4.
Jones, N. How to Stop Data Centres from Gobbling Up the World’s Electricity. Nature 2018, 561, 163–166.
- 5.
Masanet, E.; Shehabi, A.; Lei, N.; et al. Recalibrating Global Data Center Energy-Use Estimates. Science 2020, 367, 984–986.
- 6.
Luccioni, A.S.; Viguier, S.; Ligozat, A.-L. Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model. J. Mach. Learn. Res. 2023, 24, 1–15.
- 7.
Raupach, M.R.; Marland, G.; Ciais, P.; et al. Global and Regional Drivers of Accelerating CO2 Emissions. Proc. Natl. Acad. Sci. USA 2007, 104, 10288–10293. https://doi.org/10.1073/pnas.0700609104.
- 8.
Li, J.G.; Gong, F.M.; Wang, Y.B. How Does Artificial Intelligence Technology Innovation Affect Carbon Emission Reduction in Manufacturing? China Popul. Resour. Environ. 2025, 35, 14–24. (In Chinese)
- 9.
Lacoste, A.; Luccioni, A.; Schmidt, V.; et al. Quantifying the Carbon Footprint of Machine Learning. arXiv 2019, arXiv:1910.09700.
- 10.
Budennyy, S.; Lazarev, V.; Zakharenko, N.; et al. Eco2AI: Carbon Emissions Tracking of Machine Learning Models as the First Step towards Sustainable AI. arXiv 2022, arXiv:2208.00406.
- 11.
Luccioni, A.S.; Jernite, Y.; Strubell, E. Power Hungry Processing: Watts Driving the Cost of AI Deployment. arXiv 2023, arXiv:2311.16863.
- 12.
Sun, Z.Q.; Yang, R. Research on the Impact of AI Technology Innovation on Regional Carbon Emissions: Mechanism Identification and Rebound Effect. Sci. Technol. Manag. Res. 2024, 44, 168–177. (In Chinese)
- 13.
Rogers, E.M. Diffusion of Innovations, 5th ed.; Free Press: New York, NY, USA, 2003.
- 14.
Wang, X.W.; Gong, Y.L.; Xie, T.P. Mechanism and Spatial Effects of Artificial Intelligence on Provincial Carbon Emissions. Enterp. Econ. 2025, 44, 140–150. (In Chinese)
- 15.
Li, L. The Environmental Spillovers of Buyers’ Digital Transformation: Evidence from China. Technol. Forecast. Soc. Chang. 2024, 209, 123828. https://doi.org/10.1016/j.techfore.2024.123828.
- 16.
Liu, S.; Cao, Y.J. Digital Carbon Neutralization: An Empirical Study on the Dual Effects of Artificial #br# Intelligence in Promoting Industrial Carbon Reduction#br. J. Jiangsu Univ. 2025, 27, 26–39. (In Chinese)
- 17.
Mensah, J. Sustainable Development: Meaning, History, Principles, Pillars, and Implications for Human Action—Literature Review. Cogent Soc. Sci. 2019, 5, 1653531. https://doi.org/10.1080/23311886.2019.1653531.
- 18.
Patterson, D.; Gonzalez, J.; Van Der Merwe, J. The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink. Nat. Mach. Intell. 2021, 3, 875–876.
- 19.
Zhao, P.; Gao, Y.; Wu, M.; et al. How Artificial Intelligence Affects Carbon Intensity: Heterogeneous and Mediating Analyses. Environ. Dev. Sustain. 2024, 28, 2301–2325.
- 20.
Cao, Z.; Ding, Z.J. Energy Conservation and Emission Reduction Effects of Intelligent Development: Theoretical and Empirical Tests. Stat. Decis. 2023, 39, 174–178. (In Chinese)
- 21.
Kaack, L.H.; Donti, P.L.; Strubell, E.; et al. Aligning Artificial Intelligence with Climate Change Mitigation. Nat. Clim. Chang. 2022, 12, 518–527. https://doi.org/10.1038/s41558-022-01377-7.
- 22.
Schwartz, R.; Dodge, J.; Smith, N.A.; et al. Green AI. Commun. ACM 2020, 63, 54–63.
- 23.
de Vries, A. The Growing Energy Footprint of Artificial Intelligence. Joule 2023, 7, 2191–2194.
- 24.
de Vries-Gao, A. The Carbon and Water Footprints of Data Centers and What This Could Mean for Artificial Intelligence. Patterns 2026, 7, 101430. https://doi.org/10.1016/j.patter.2025.101430.
- 25.
Vercellino, R.; Willard, J.; Campos, G.; et al. Measurement of Generative AI Workload Power Profiles for Whole-Facility Data Center Infrastructure Planning. arXiv 2026, arXiv:2604.07345.
- 26.
Elsworth, C.; Huang, K.; Patterson, D.; et al. Measuring the Environmental Impact of Delivering AI at Scale. arXiv 2025, arXiv:2508.15734.
- 27.
Shehabi, A.; Smith, S.; Sartor, D.; et al. United States Data Center Energy Usage Report; Lawrence Berkeley National Laboratory: Berkeley, CA, USA, 2016.
- 28.
National Development and Reform Commission; Cyberspace Administration of China; Ministry of Industry and Information Technology; National Energy Administration. Implementation Plan for the Computing Hubs of the National Integrated Big Data Center Collaborative Innovation System; FGGT [2021] No. 709; Beijing, China, 2021. Available online: https://www.ndrc.gov.cn/xxgk/zcfb/tz/202105/t20210526_1280838.html (accessed on 25 June 2026).
- 29.
Porter, M.E.; van der Linde, C. Toward a New Conception of the Environment-Competitiveness Relationship. J. Econ. Perspect. 1995, 9, 97–118.
This work is licensed under a Creative Commons Attribution 4.0 International License.
Suite 4002 Level 4, 447 Collins Street, Melbourne, Victoria 3000, Australia
General Inquiries: info@sciltp.com