There is a growing awareness of the substantial environmental costs of large language models (LLMs), but discussing the sustainability of LLMs only in terms of CO2 emissions is not enough. This Comment emphasizes the need to take into account the social and ecological costs and benefits of LLMs as well.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 /Â 30Â days
cancel any time
Subscribe to this journal
Receive 12 digital issues and online access to articles
$119.00 per year
only $9.92 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
References
Strubell, E., Ganesh, A. & McCallum, A. In Proc. 57th Annual Meeting of the Association for Computational Linguistics 3645â3650 (ACL, 2019).
Luccioni, A. S., Viguier, S. & Ligozat, A.-L. J. Mach. Learn. Res. 24, 253 (2022).
Bender, E., Gebru, T., McMillan-Major, A. & Shmitchell, S. In Proc. 2021 ACM Conference on Fairness, Accountability, and Transparency (FAcct 2021) 610â623 (ACM, 2021).
Coeckelbergh, M. AI Ethics 1, 67â72 (2021).
Crownhart, C. MIT Technol. Rev. https://go.nature.com/40eKtAz (2024).
World Commission on Environment and Development (WCED). Our Common Future (Oxford Univ. Press, 1987).
Kubiszewski, I. et al. Ecol. Econom. 93, 57â68 (2013).
European Commission. Ethics Guidelines for Trustworthy AI https://go.nature.com/3WiIMke (2019).
Mittelstadt, B. Nature Mach. Intell. 1, 501â507 (2019).
Bolte, L., Vandemeulebroucke, T. & Van Wynsberghe, A. Sustainability 14, 4472 (2022).
Willet, J. in Environmental Justice (eds Fogel, S. et al.) Ch. 7 (Routledge, (2016).
Haider, J., Rödl, M. & Joosse, S. Information Res. https://doi.org/10.47989/colis2224 (2022).
Future of Life Institute. Pause Giant AI Experiments: An Open Letter https://go.nature.com/44Apd9V (2023).
Collingridge, D. The Social Control of Technology (Pinter, 1982).
Metz, C. & Schmidt, G. The New York Times https://go.nature.com/40dN2my (23 May 2024).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Rights and permissions
About this article
Cite this article
Bossert, L.N., Loh, W. Why the carbon footprint of generative large language models alone will not help us assess their sustainability. Nat Mach Intell (2025). https://doi.org/10.1038/s42256-025-00979-y
Published:
DOI: https://doi.org/10.1038/s42256-025-00979-y
