Using AI models requires large amounts of water
Is it true that using AI models requires large amounts of water?
Yes. Multiple studies and data center assessments confirm that training and operating artificial intelligence models requires substantial quantities of freshwater for cooling and power generation.
Verdict: SUPPORTED · 6 sources for, 0 against · Last checked
The claim is that using AI models requires large amounts of water. Multiple papers (such as [2], [3], [4], and [7]) explicitly quantify the substantial water footprints, freshwater consumption for cooling, and resource demands associated with training and running AI inference queries. There are no papers refuting this evidence, leading to a verdict of SUPPORTED.
Get citation
Ren, S., Tomlinson, B., Black, R. W., & Torrance, A. (2024). Reconciling the contrasting narratives on the environmental impact of large language models. doi.org. https://doi.org/10.1038/s41598-024-76682-6
See more details
Get citation
Moore, H., Qi, S., Hogade, N., Milojicic, D., Bash, C. E., & Pasricha, S. (2025). Sustainable Carbon-Aware and Water-Efficient LLM Scheduling in Geo-Distributed Cloud Datacenters. doi.org. https://doi.org/10.1145/3716368.3735301
Get citation
Goff, T. L. (2024). Recommendations for public action towards sustainable generative AI systems. doi.org. https://doi.org/10.48550/arXiv.2402.01646
Get citation
A, d. V. (2026). The carbon and water footprints of data centers and what this could mean for artificial intelligence.. doi.org. https://doi.org/10.1016/j.patter.2025.101430
Get citation
Z, B. (2026). The water footprint of artificial intelligence: Emerging solutions and governance imperatives.. doi.org. https://doi.org/10.1016/j.watres.2026.125866
Get citation
E, S. (2026). [Is it possible to envision a sustainable artificial intelligence?]. doi.org. https://doi.org/10.1701/4649.46620
- Is it true that all large language models hallucinate? PARTIAL LEAN
- Is it true that training and running large language models consumes significant water resources? SUPPORTED
- Is it true that ChatGPT is powered by a large language model? SUPPORTED
- Is it true that large language models can run on consumer laptops? PARTIAL LEAN
- Is it true that trees are the exclusive source of large amounts of atmospheric oxygen? REFUTED
- Large language models can be made to stop hallucinating entirely using current techniques such as retrieval augmentation REFUTED
What does the evidence say?
The claim is that using AI models requires large amounts of water. Multiple papers (such as [2], [3], [4], and [7]) explicitly quantify the substantial water footprints, freshwater consumption for cooling, and resource demands associated with training and running AI inference queries. There are no papers refuting this evidence, leading to a verdict of SUPPORTED.
Receipt History
Challenge the receipt
Citation formatting by citeproc-js (Frank Bennett) and the Citation Style Language project. Source and licenses.
Terms · Privacy · How verdicts work