There is great interest in how the growth of artificial intelligence and machine learning may affect global GHG emissions. However, such emissions impacts remain uncertain, owing in part to the diverse mechanisms through which they occur, posing difficulties for measurement and forecasting. Here we introduce a systematic framework for describing the effects of machine learning (ML) on GHG emissions, encompassing three categories: computing-related impacts, immediate impacts of applying ML and system-level impacts. Using this framework, we identify priorities for impact assessment and scenario analysis, and suggest policy levers for better understanding and shaping the effects of ML on climate change mitigation.
Your institute does not have access to this article
Subscribe to Nature+
Get immediate online access to the entire Nature family of 50+ journals
Subscribe to Journal
Get full journal access for 1 year
only $8.25 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
Zhang, D. et al. Artificial Intelligence Index Report 2021 (AI Index Steering Committee, Human-Centered AI Institute, 2021).
Digital Technology and the Planet: Harnessing Computing to Achieve Net Zero (Royal Society, 2020); https://royalsociety.org/-/media/policy/projects/digital-technology-and-the-planet/digital-technology-and-the-planet-report.pdf
Kaack, L. H., Donti, P. L., Strubell, E. & Rolnick, D. Artificial Intelligence and Climate Change: Opportunities, Considerations, and Policy Levers to Align AI with Climate Change Goals (Heinrich-Böll-Stiftung, 2020); https://eu.boell.org/en/2020/12/03/artificial-intelligence-and-climate-change
Harnessing Artificial Intelligence to Accelerate the Energy Transition (World Economic Forum, 2021); https://www.weforum.org/whitepapers/harnessing-artificial-intelligence-to-accelerate-the-energy-transition
Berkhout, F. & Hertin, J. De-materialising and re-materialising: digital technologies and the environment. Futures 36, 903–920 (2004).
Hilty, L. M. & Aebischer, B. in ICT Innovations for Sustainability (eds Hilty, L. M. & Aebischer, B) 3–36 (Springer, 2015).
Rolnick, D. et al. Tackling climate change with machine learning. ACM Comput. Surv. https://doi.org/10.1145/3485128 (2022).
Oil in the Cloud: How Tech Companies are Helping Big Oil Profit from Climate Destruction (Greenpeace, 2019); https://www.greenpeace.org/usa/reports/oil-in-the-cloud/
Dobbe, R. & Whittaker, M. AI and climate change: how they’re connected, and what we can do about it. Medium https://medium.com/@AINowInstitute/ai-and-climate-change-how-theyre-connected-and-what-we-can-do-about-it-6aa8d0f5b32c (2019).
Strubell, E., Ganesh, A. & McCallum, A. Energy and policy considerations for deep learning in NLP. In Proc. 57th Annual Meeting of the Association for Computational Linguistics 3645–3650 (Association for Computational Linguistics, 2019); https://doi.org/10.18653/v1/P19-1355
Schwartz, R., Dodge, J., Smith, N. A. & Etzioni, O. Green AI. Commun. ACM 63, 54–63 (2020).
Dauvergne, P. Is artificial intelligence greening global supply chains? Exposing the political economy of environmental costs. Rev. Int. Polit. Econ. https://doi.org/10.1080/09692290.2020.1814381 (2020).
Coeckelbergh, M. AI for climate: freedom, justice, and other ethical and political challenges. AI Ethics 1, 67–72 (2021).
Gunther, H. & Rose, J. Governing AI: the importance of environmentally sustainable and equitable innovation. Environ. Law Rep. 50, 10888 (2020).
Stein, A. L. Artificial intelligence and climate change. Yale J. Reg. 37, 890–939 (2020).
Cowls, J., Tsamados, A., Taddeo, M. & Floridi, L. The AI gambit-leveraging artificial intelligence to combat climate change: opportunities, challenges, and recommendations. SSRN https://ssrn.com/abstract=3804983 (2021).
FAccT ’21: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (ACM, 2021).
Recommendation on the Ethics of Artificial Intelligence (UNESCO, 2021); https://unesdoc.unesco.org/ark:/48223/pf0000380455
Areas for Future Action in the Responsible AI Ecosystem (The Future Society, GPAI Responsible Development, Use and Governance of AI Working Group & CEIMIA, 2020); https://www.gpai.ai/projects/responsible-ai/areas-for-future-action-in-responsible-ai.pdf
Horner, N. C., Shehabi, A. & Azevedo, I. L. Known unknowns: indirect energy effects of information and communication technology. Environ. Res. Lett. 11, 103001 (2016).
Bieser, J. & Hilty, L. Indirect effects of the digital transformation on environmental sustainability: methodological challenges in assessing the greenhouse gas abatement potential of ICT. In 5th International Conference on Information and Communication Technology for Sustainability 68–81 (EasyChair, 2018); https://doi.org/10.29007/lx7q
Pohl, J., Hilty, L. M. & Finkbeiner, M. How LCA contributes to the environmental assessment of higher order effects of ICT application: a review of different approaches. J. Clean. Prod. 219, 698–712 (2019).
Digitalization & Energy (OECD/IEA, 2017); https://www.iea.org/reports/digitalisation-and-energy
Sivaram, V. et al. Digital Decarbonization Promoting Digital Innovations to Advance Clean Energy Systems (Council on Foreign Relations, 2018); https://www.cfr.org/report/digital-decarbonization
Wilson, C., Kerr, L., Sprei, F., Vrain, E. & Wilson, M. Potential climate benefits of digital consumer innovations. Annu. Rev. Environ. Resour. 45, 113–144 (2020).
Canziani, A., Paszke, A. & Culurciello, E. An analysis of deep neural network models for practical applications. Preprint at https://arxiv.org/abs/1605.07678 (2017).
AI and compute. OpenAI https://openai.com/blog/ai-and-compute (16 May 2018).
Bommasani, R. et al. On the opportunities and risks of foundation models. Preprint at https://arxiv.org/abs/2108.07258 (2021).
Joulin, A., Grave, E., Bojanowski, P. & Mikolov, T. Bag of tricks for efficient text classification. In Proc. 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 2 427–431 (Association for Computational Linguistics, 2017); https://www.aclweb.org/anthology/E17-2068
Hazelwood, K. et al. Applied machine learning at Facebook: a datacenter infrastructure perspective. In 2018 IEEE International Symposium on High Performance Computer Architecture 620–629 (IEEE, 2018); https://doi.org/10.1109/HPCA.2018.00059
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A. & Chen, L.-C. MobileNetV2: inverted residuals and linear bottlenecks. In IEEE Conference on Computer Vision and Pattern Recognition (IEEE, 2018).
Turovsky, B. Ten years of Google Translate. The Keyword https://blog.google/products/translate/ten-years-of-google-translate (2016).
Wu, C. J. et al. Sustainable AI: environmental implications, challenges and opportunities. In Proc. Machine Learning and Systems 4 795–813 (MLSys, 2022).
Jiang, A. H. et al. Accelerating deep learning by focusing on the biggest losers. Preprint at https://arxiv.org/abs/1910.00762 (2019).
He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition 770–778 (IEEE, 2016); https://doi.org/10.1109/CVPR.2016.90
Albanie, S. ConvNet-Burden: estimates of memory consumption and FLOP counts for various convolutional neural networks. GitHub https://github.com/albanie/convnet-burden (2019).
Bender, E. M., Gebru, T., McMillan-Major, A. & Shmitchell, S. On the dangers of stochastic parrots: can language models be too big? In Proc. 2021 ACM Conference on Fairness, Accountability, and Transparency 610–623 (ACM, 2021).
Gupta, A., Lanteigne, C. & Kingsley, S. SECure: a social and environmental certificate for AI Systems. Preprint at https://arxiv.org/abs/2006.06217 (2020).
Tomašev, N. et al. AI for social good: unlocking the opportunity for positive impact. Nat. Commun. 11, 2468 (2020).
Henderson, P. et al. Towards the systematic reporting of the energy and carbon footprints of machine learning. J. Mach. Learn. Res. 21, 1–43 (2020).
Schmidt, V. et al. CodeCarbon: estimate and track carbon emissions from machine learning computing. Zenodo https://doi.org/10.5281/zenodo.4658424 (2021).
Anthony, L. F. W., Kanding, B. & Selvan, R. Carbontracker: tracking and predicting the carbon footprint of training deep learning models. Preprint at https://arxiv.org/abs/2007.03051 (2020).
Cai, E., Juan, D., Stamoulis, D. & Marculescu, D. NeuralPower: predict and deploy energy-efficient convolutional neural networks. In 9th Asian Conference on Machine Learning (ACML, 2017).
Dodge, J., Gururangan, S., Card, D., Schwartz, R. & Smith, N. A. Show your work: improved reporting of experimental results. In Proc. 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing 2185–2194 (Association for Computational Linguistics, 2019); https://doi.org/10.18653/v1/D19-1224
Mattson, P. et al. (eds) Proc. Machine Learning and Systems 2 336–349 (MLSys, 2020).
Reddi, V. J. et al. MLPerf inference benchmark. In 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture 446–459 (IEEE, 2020); https://doi.org/10.1109/ISCA45697.2020.00045
Hinton, G., Vinyals, O. & Dean, J. Distilling the knowledge in a neural network. In NeurIPS Deep Learning Workshop (NeurIPS, 2014).
Schaul, T., Quan, J., Antonoglou, I. & Silver, D. Prioritized experience replay. In International Conference on Learning Representations (ICLR, 2016).
Hubara, I., Courbariaux, M., Soudry, D., El-Yaniv, R. & Bengio, Y. Quantized neural networks: training neural networks with low precision weights and activations. J. Mach. Learn. Res. 18, 1–30 (2018).
Pfeiffer, J. et al. Adapterhub: a framework for adapting transformers. In Proc. 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations 46–54 (Association for Computational Linguistics, 2020).
Cai, H., Gan, C., Wang, T., Zhang, Z. & Han, S. Once-for-all: train one network and specialize it for efficient deployment. In International Conference on Learning Representations (ICLR, 2020).
Lepikhin, D. et al. GShard: scaling giant models with conditional computation and automatic sharding. In International Conference on Learning Representations (ICLR, 2021).
Hooker, S., Moorosi, N., Clark, G., Bengio, S. & Denton, E. Characterizing and mitigating bias in compact models. In ICML Workshop on Human Interpretability in Machine Learning (ICML, 2020).
Greenhouse Gas Emissions Trajectories for the Information and Communication Technology Sector Compatible with the UNFCCC Paris Agreement (International Telecommunication Union, 2020); http://handle.itu.int/11.1002/1000/14084
Malmodin, J. & Lundén, D. The energy and carbon footprint of the global ICT and E&M sectors 2010–2015. Sustainability 10, 3027 (2018).
Masanet, E., Shehabi, A., Lei, N., Smith, S. & Koomey, J. Recalibrating global data center energy-use estimates. Science 367, 984–986 (2020).
Data Centres and Data Transmission Networks (International Energy Agency, 2021); https://www.iea.org/reports/data-centres-and-data-transmission-networks
Montevecchi, F., Stickler, T., Hintemann, R. & Hinterholzer, S. Energy-Efficient Cloud Computing Technologies and Policies for an Eco-Friendly Cloud Market (2020); https://ec.europa.eu/newsroom/dae/document.cfm?doc_id=71330
Cisco Global Cloud Index: Forecast and Methodology, 2016–2021 (Cisco, 2018); https://www.cisco.com/c/en/us/solutions/collateral/service-provider/global-cloud-index-gci/white-paper-c11-738085.pdf
Compton, C. Cisco’s Global Cloud Index study: acceleration of the multicloud era. Cisco Blogs https://blogs.cisco.com/news/acceleration-of-multicloud-era (2018).
Wu, C. et al. Machine learning at Facebook: understanding inference at the edge. In 2019 IEEE International Symposium on High Performance Computer Architecture 331–344 (IEEE, 2019).
Koomey, J., Berard, S., Sanchez, M. & Wong, H. Implications of historical trends in the electrical efficiency of computing. IEEE Ann. Hist. Comput. 33, 46–54 (2010).
Koomey, J. & Naffziger, S. Moore's law might be slowing down, but not energy efficiency. IEEE Spectrum 52, 35 (2015).
Facebook Sustainability Data 2020 (Facebook, 2021); https://sustainability.fb.com/wp-content/uploads/2021/06/2020_FB_Sustainability-Data.pdf
Naumov, M. et al. Deep learning training in Facebook data centers: design of scale-up and scale-out systems. Preprint at https://arxiv.org/abs/2003.09518 (2020).
Park, J. et al. Deep learning inference in Facebook data centers: characterization, performance optimizations and hardware implications. Preprint at https://arxiv.org/abs/1811.09886 (2018).
Shehabi, A. et al. United States Data Center Energy Usage Report (Lawrence Berkeley National Laboratory, 2016); https://eta.lbl.gov/publications/united-states-data-center-energy
Jouppi, N. P. et al. In-datacenter performance analysis of a tensor processing unit. In Proc. 44th Annual International Symposium on Computer Architecture 1–12 (ACM, 2017).
Radovanovic, A. Our data centers now work harder when the sun shines and wind blows. The Keyword https://blog.google/inside-google/infrastructure/data-centers-work-harder-sun-shines-wind-blows (2020).
Whitehead, B., Andrews, D. & Shah, A. The life cycle assessment of a UK data centre. Int. J. Life Cycle Assess. 20, 332–349 (2015).
Masanet, E., Shehabi, A. & Koomey, J. Characteristics of low-carbon data centres. Nat. Clim. Change 3, 627–630 (2013).
Hischier, R., Coroama, V. C., Schien, D. & Achachlouei, M. A. in ICT Innovations for Sustainability (eds Hilty, L. M. & Aebischer, B.) 171–189 (Springer, 2015).
André Barroso, L., Clidaras, J. & Hölzle, U. The datacenter as a computer: an introduction to the design of warehouse-scale machines. Synth. Lect. Comput. Arch. 8, 1–154 (2013).
Gupta, U. et al. Chasing carbon: the elusive environmental footprint of computing. In IEEE International Symposium on High-Performance Computer Architecture 854–867 (IEEE, 2021).
Finer, M. et al. Combating deforestation: from satellite to intervention. Science 360, 1303–1305 (2018).
Kulp, S. A. & Strauss, B. H. New elevation data triple estimates of global vulnerability to sea-level rise and coastal flooding. Nat. Commun. 10, 4844 (2019).
Friederich, D., Kaack, L. H., Luccioni, A. & Steffen, B. Automated identification of climate risk disclosures in annual corporate reports. Preprint at https://arxiv.org/abs/2108.01415 (2021).
Liu, Y., Guo, B., Zou, X., Li, Y. & Shi, S. Machine learning assisted materials design and discovery for rechargeable batteries. Energy Storage Mater. 31, 434–450 (2020).
Ahmed, R., Sreeram, V., Mishra, Y. & Arif, M. D. A review and evaluation of the state-of-the-art in PV solar power forecasting: techniques and optimization. Renew. Sustain. Energy Rev. 124, 109792 (2020).
You, J., Li, X., Low, M., Lobell, D. & Ermon, S. Deep Gaussian process for crop yield prediction based on remote sensing data. In Thirty-First AAAI Conference on Artificial Intelligence 4559–4565 (ACM, 2017).
Toqué, F., Khouadjia, M., Come, E., Trepanier, M. & Oukhellou, L. Short & long term forecasting of multimodal transport passenger flows with machine learning methods. In 2017 IEEE 20th International Conference on Intelligent Transportation Systems 560–566 (IEEE, 2017).
Evans, R. & Gao, J. DeepMind AI reduces Google data centre cooling bill by 40%. DeepMind https://www.deepmind.com/blog/deepmind-ai-reduces-google-data-centre-cooling-bill-by-40 (2016).
Roman, N. D., Bre, F., Fachinotti, V. D. & Lamberts, R. Application and characterization of metamodels based on artificial neural networks for building performance simulation: a systematic review. Energy Build. 217, 109972 (2020).
Irrgang, C. et al. Towards neural Earth system modelling by integrating artificial intelligence in Earth system science. Nat. Mach. Intell. 3, 667–674 (2021).
Jenssen, R. et al. Automatic autonomous vision-based power line inspection: a review of current status and the potential role of deep learning. Int. J. Electr. Power Energy Syst. 99, 107–120 (2018).
Rudin, C. et al. Interpretable machine learning: fundamental principles and 10 grand challenges. Stat. Surv. 16, 1–85 (2022).
Ghahramani, Z. Probabilistic machine learning and artificial intelligence. Nature 521, 452–459 (2015).
Willard, J., Jia, X., Xu, S., Steinbach, M. & Kumar, V. Integrating physics-based modeling with machine learning: a survey. Preprint at https://arxiv.org/abs/2003.04919 (2020).
Zhuang, F. et al. A comprehensive survey on transfer learning. Proc. IEEE 109, 43–76 (2020).
Adams-Progar, A., Fink, G.A.,Walker, E. & Llewellyn, D. in Security and Privacy in Cyber‐Physical Systems: Foundations, Principles and Applications (eds Song, H. et al.) Ch. 18 (Wiley, 2017); https://doi.org/10.1002/9781119226079.ch18
Charles, H. et al. Meat consumption, health, and the environment. Science 361, eaam5324 (2018).
Herweijer, C., Combes, B. & Gillham, J. How AI Can Enable a Sustainable Future (Microsoft & PWC, 2018); https://www.pwc.co.uk/services/sustainability-climate-change/insights/how-ai-future-can-enable-sustainable-future.html
Climate AI: How Artificial Intelligence Can Power Your Climate Action Strategy (Capgemini, 2020); https://www.capgemini.com/research/climate-ai/
Degot, C., Duranton, S., Frédeau, M. & Hutchinson, R. Reduce carbon and costs with the power of AI. BCG https://www.bcg.com/en-us/publications/2021/ai-to-reduce-carbon-emissions (2021).
Azevedo, I. M. L. Consumer end-use energy efficiency and rebound effects. Annu. Rev. Environ. Resour. 39, 393–418 (2014).
Lange, S., Pohl, J. & Santarius, T. Digitalization and energy consumption. Does ICT reduce energy demand? Ecol. Econ. 176, 106760 (2020).
Anderson, J. M. et al. Autonomous Vehicle Technology: A Guide for Policymakers (RAND Corporation, 2016); https://doi.org/10.7249/RR443-2
Creutzig, F. et al. Leveraging digitalization for sustainability in urban transport. Glob. Sustain. 2, e14 (2019).
Wadud, Z., MacKenzie, D. & Leiby, P. Help or hindrance? The travel, energy and carbon impacts of highly automated vehicles. Transport. Res. A 86, 1–18 (2016).
Chase, N., Maples, J. & Schipper, M. Autonomous Vehicles: Uncertainties and Energy Implications (EIA, 2018); https://www.eia.gov/outlooks/aeo/av.php
Arthur, W. B. Competing technologies, increasing returns, and lock-in by historical events. Econ. J. 99, 116–131 (1989).
Cox, E., Royston, S. & Selby, J. Impact of Non-Energy Policies on Energy Systems (UK Energy Research Centre, 2016); https://ukerc.ac.uk/publications/impact-of-non-energy-policies-on-energy-systems/
Stilgoe, J., Owen, R. & Macnaghten, P. Developing a framework for responsible innovation. Res. Policy 42, 1568–1580 (2013).
Jirotka, M., Grimpe, B., Stahl, B., Eden, G. & Hartswood, M. Responsible research and innovation in the digital age. Commun. ACM 60, 62–68 (2017).
Itten, R. et al. Digital transformation-life cycle assessment of digital services, multifunctional devices and cloud computing. Int. J. Life Cycle Assess. 25, 2093–2098 (2020).
Coroamă, V. C., Bergmark, P., Höjer, M. & Malmodin, J. A methodology for assessing the environmental effects induced by ICT services: part I. Single services. In Proc. 7th International Conference on ICT for Sustainability 36–45 (ACM, 2020).
Bergmark, P., Coroamă, V. C., Höjer, M. & Donovan, C. A Methodology for assessing the environmental effects induced by ICT services: part I. Multiple services and companies. In Proc. 7th International Conference on ICT for Sustainability 46–55 (ACM, 2020).
Haataja, M. & Bryson, J. J. What costs should we expect from the EU’s AI Act? Preprint at SocArXiv https://osf.io/preprints/socarxiv/8nzb4 (2021).
Mytton, D. Hiding greenhouse gas emissions in the cloud. Nat. Clim. Change 10, 701–701 (2020).
Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and Amending Certain Union Legislative Acts (European Commission, 2021); https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021PC0206
Hilbert, M. Big data for development: a review of promises and challenges. Dev. Policy Rev. 34, 135–174 (2016).
Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K. & Galstyan, A. A survey on bias and fairness in machine learning. in ACM Computing Surveys Vol. 54, 1–35 (ACM, 2021).
Bondi, E., Xu, L., Acosta-Navas, D. & Killian, J. A. Envisioning communities: a participatory approach towards AI for social good. In Proc. 2021 AAAI/ACM Conference on AI, Ethics, and Society 425–436 (ACM, 2021).
Pinch, T. J. & Bijker, W. E. The social construction of facts and artefacts: or how the sociology of science and the sociology of technology might benefit each other. Soc. Stud. Sci. 14, 399–441 (1984).
Klein, H. K. & Lee Kleinman, D. The social construction of technology: structural considerations. Sci. Technol. Human Values 27, 28–52 (2002).
We thank the Environmental Law Institute and Alfred P. Sloan Foundation for their support of this work. P.L.D. was also supported by a US Department of Energy Computational Science Graduate Fellowship (DE-FG02-97ER25308), the Center for Climate and Energy Decision Making through a cooperative agreement between the US National Science Foundation and Carnegie Mellon University (grant number SES-00949710) and the Siebel Scholars programme. D.R. was supported by the Canada CIFAR AI Chairs programme. We are grateful to W.Y. for assistance with the figures. We thank P. Bergmark, V. Coroama, D. Daniels, J. Dunietz, L. Klaaβen, S. Luccioni, J. Malmodin, D. Rejeski, C. Samaras, N. Schmid, T. S. Schmidt, M. Scheutz, S. Sewerin, B. Steffen and M. Voss for their input and comments on the manuscript. Opinions are G.K.’s own and do not reflect those of the OECD, the IEA or their member countries.
The authors declare no competing interests.
Peer review information
Nature Climate Change thanks Lorenz Hilty, Christopher Irrgang and Yang Yu for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Kaack, L.H., Donti, P.L., Strubell, E. et al. Aligning artificial intelligence with climate change mitigation. Nat. Clim. Chang. 12, 518–527 (2022). https://doi.org/10.1038/s41558-022-01377-7