Pith. sign in

REVIEW 4 major objections 6 minor 37 references

Misinformation by Omission: The Need for More Environmental Transparency in AI

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read AI models have grown less transparent about their environmental costs since 2022, and the figures that fill the information gap are misinformation grown from decontextualized estimates.

desk verdict Useful myth-tracing and a plausible but under-audited transparency trend; the Indirect Disclosure category needs a sensitivity check before I'd trust the headline. read the letter →

arxiv 2506.15572 v1 pith:ZFWAIRKS submitted 2025-06-18 cs.CY

Sasha Luccioni , Boris Gamazaychikov , Theo Alves da Costa , Emma Strubell This is my paper
classification cs.CY
keywords environmentaltransparencyAIenergyconsumptiongreenhousegasemissionsmodeldisclosuremisinformationEpochnotablemodelsLLMinference
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that the AI industry is becoming less transparent about the environmental costs of its models just as those costs are growing. Combining Epoch AI's list of 754 notable models from 2010 to early 2025, the authors find that direct disclosure of energy use or greenhouse-gas emissions peaked in 2022 and then declined, so that by the first quarter of 2025 most notable models disclosed nothing usable. The authors then trace three widely repeated numbers, the "five cars" estimate for training emissions, the claim that a ChatGPT query uses ten times the energy of a Google search, and the claim that AI can cut global emissions by 5 to 10 percent, to show how best-effort, decontextualized estimates become misinformation. If they are right, the paper establishes that the environmental transparency problem is not just missing data but active misunderstanding, and that users and policymakers are making decisions from figures that do not represent actual impacts.

What carries the argument

The central machinery is a three-category transparency classification applied to a curated dataset: Direct Disclosure (the developer reports energy or GHG numbers), Indirect Disclosure (training compute or model weights are released, allowing outsiders to estimate impacts), and No Disclosure (neither is available). The same classification applied to OpenRouter's top-20 monthly LLM usage list produces the headline usage split of 84 percent no disclosure, 14 percent indirect, and 2 percent direct. This coding is what turns the claim of declining transparency into a measurable trend rather than an anecdote, and the provenance tracing of the "five cars", "3 Wh", and "5 to 10 percent" figures is the second mechanism showing how isolated estimates become public misinformation.

What would settle it

Independently re-code a random sample of 100 models released in 2022 and 100 released in the first quarter of 2025 for direct, indirect, or no environmental disclosure using a pre-registered codebook; if the share of no-disclosure models is not higher in the later period, the claimed decline would be an artifact of the dataset curation or the manual coding. A complementary check would count how many of the top-20 OpenRouter models' publishers release energy or GHG figures in their model cards and compare the token-weighted disclosure share with the paper's figures.

Watch

Extended reading notes

Core claim

The paper's central discovery is a documented inversion: methodological tools for measuring AI's environmental impact have improved since 2019, yet industry disclosure has gone backwards. Classifying every model in Epoch AI's Notable AI Models dataset, 754 models from 2010 through the first quarter of 2025, into Direct Disclosure, Indirect Disclosure, or No Disclosure, the authors report that direct disclosure of energy or GHG data peaked in 2022 at 10 percent of notable models, and that by Q1 2025 the majority of notable models again fell into No Disclosure. A complementary snapshot of OpenRouter traffic in May 2025 shows that 84 percent of LLM token usage flowed through no-disclosure models, 14 percent through indirect-disclosure models, and only 2 percent through direct-disclosure models. The authors also show that the three most widespread quantitative claims about AI's climate footprint each trace to a single qualified or speculative estimate that was stripped of its context in repeated media coverage, with 75 percent of 100 sampled news articles relaying energy figures for ChatGPT queries without citing sources or expressing uncertainty.

Load-bearing premise

The load-bearing premise is that the authors' classification of 754 curated models into three transparency buckets faithfully represents the AI industry's environmental disclosure, rather than an artifact of which models Epoch AI selected or how the authors coded them.

Editorial extensions

If this is right

  • Users of the most-used LLM APIs cannot estimate the energy or carbon cost of the queries they make, so procurement decisions based on efficiency are effectively impossible.
  • Policymakers relying on "five cars" or "ten times a Google search" as if they were measured facts are likely to misallocate effort in climate regulations.
  • Because direct disclosures peaked in 2022 and then fell, attempts to regulate AI environmental impacts cannot rely on industry self-reporting in its current form.
  • The evidence that actual pretraining emissions for Gemma and Llama 3 exceed the "five cars" figure by roughly 4 times and 40 times shows that even the widely criticized estimate understates today's largest training runs.
  • Standardized, verifiable reporting frameworks and procurement requirements are the direct corollary of the finding that omission creates misinformation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An extension the authors leave implicit is that the same three-category coding could be applied to fine-tuning and API-serving layers rather than only base models, which would likely push the no-disclosure share even higher because most fine-tuning services do not report energy use.
  • A testable extension is a longitudinal audit of model cards on major hosting platforms, which could verify whether the post-2022 decline is driven by proprietary labs or by open-weight releases increasingly choosing not to disclose.
  • If mandatory reporting under the EU's sustainability disclosure rules is enforced, the paper's framework predicts that transparency will improve among labs that serve EU customers, providing a natural experiment for the claim that omission is a choice rather than a technical necessity.
  • The paper implies that the "misinformation by omission" mechanism generalizes beyond these three myths, so fact-checking infrastructure for AI-energy claims would be a low-cost mitigation even before formal standards arrive.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This paper examines environmental transparency in AI model development and deployment. Using Epoch AI's Notable AI Models dataset (754 models from 2010 to Q1 2025), the authors classify each model into three transparency categories: Direct Disclosure (energy/GHG emissions reported), Indirect Disclosure (training compute data or open weights enabling external estimates), and No Disclosure. They report that transparency improved through 2022, then reversed, with the majority of notable models in early 2025 falling into the no-disclosure category. An analysis of OpenRouter traffic for May 2025 finds that 84% of LLM token usage flows through no-disclosure models. The paper then investigates two widely circulated claims—that training a model emits as much CO2 as five cars over their lifetime, and that a ChatGPT query uses ten times more energy than a Google search—and traces their origins and misrepresentation. The authors also present a media analysis of 100 articles covering ChatGPT energy consumption, finding that 75% relayed energy estimates without uncertainty or sourcing. The paper concludes with recommendations for measurement, standards, verification, and policy.

Significance. If the documented trend is robust, the paper makes an important contribution by showing that environmental opacity in AI is worsening as impacts grow, and it identifies concrete misinformation pathways. The paper's careful treatment of the 'five cars' estimate—distinguishing the upper-bound neural architecture search scenario from typical training workloads—is a genuine strength, as is the provenance tracing of the 'ten times a Google search' claim. The analysis produces falsifiable claims: the transparency trend can be re-tested with alternative coding schemes, and the media analysis percentages can be audited. The recommendation section provides actionable, if high-level, pathways. However, the paper's own empirical contributions rest on a category definition that conflates open-weights licensing with environmental disclosure, and on manual coding without released data or reliability checks; these need to be addressed before the central claim can be considered established.

major comments (4)
  1. [Environmental Transparency Trends (definition of Indirect Disclosure)] The definition of Indirect Disclosure—'Developers provided training compute data or released their model weights, allowing external estimates of training or inference impacts'—conflates model accessibility with environmental transparency. Under this coding, a model with open weights and no environmental information whatsoever is classified as more transparent than a closed model that reports training compute or emissions. Because the post-2022 reversal is attributed to 'the introduction of increasingly commercial and proprietary models,' the headline trend may simply track the shift from open-weights research releases to closed APIs rather than a decline in environmental reporting. Please re-analyze the data with a stricter coding scheme that separates release of weights from release of compute/emissions data (e.g., treating open weights alone as no disclosure when no compute or energy data are provided), and report the trend under both codings. If the decline attenuates or disappears, the central claim must be revised.
  2. [Environmental Transparency Trends / Methods (manual coding)] The classification of all 754 Epoch models and the coding of the 100 news articles are performed by the authors without a published codebook, released raw data, or inter-rater reliability checks. The quantitative headline claims—such as 84% of OpenRouter usage through no-disclosure models, and the period-by-period percentages in Figure 1—cannot be independently verified. Please make the coded dataset and codebook available, and provide reliability statistics (e.g., dual coding of a random sample) or, at minimum, a sensitivity analysis showing that the conclusions are robust to plausible coding disagreements.
  3. [Investigating the Urban Legends (media analysis)] The media analysis described in the section 'Investigating the Urban Legends' reports that 75% of 100 articles relayed energy estimates without uncertainty or sourcing, but the methodology is under-specified: the sample was taken from a Google News search for 'ChatGPT energy consumption' as of April 11, 2025, but the text does not state how the 100 articles were selected from the results, what inclusion/exclusion criteria were used, how the coding categories were defined, or whether coding was performed by more than one person. These details are needed to assess whether the 75% figure is reliable and generalizable. Please add a methods paragraph describing article selection, coding rules, and coder agreement.
  4. [OpenRouter analysis (Figure 2)] The paper states that '84% of LLM usage is through models with no disclosure' based on OpenRouter's top 20 models in May 2025, but the text does not clarify whether this 84% is computed only over the top 20 or over total tokens on the platform. If OpenRouter data are limited to the top 20, the claim should be qualified as 'the share among the top 20 most-used models' and the paper should discuss whether the top 20 dominate overall token volume. Without this caveat, the statistic overstates the representativeness of the snapshot.
minor comments (6)
  1. [Environmental Transparency Trends] The text contains a typo: 'This period includes the the work of Strubell et al.' — the duplicate 'the' should be removed.
  2. [Investigating the Urban Legends] The sentence 'This remark was used was the basis of an estimate published in October 2023' contains an extra 'was'; it should read 'This remark was used as the basis...'.
  3. [Appendix Table 1] Several entries in Table 1 are listed as '?' (e.g., Gemma 2B+9B energy and GHG, Llama 3 70B energy), and the 'Max/Min Variance' row is not defined; please add a note explaining the table's conventions and what the variance row represents.
  4. [Introduction / How to improve environmental impact disclosures] The authors recommend the AI Energy Score project without a conflict-of-interest statement; given that two authors are directly involved in that project, a formal disclosure should be included.
  5. [Figure 1] The caption of Figure 1 does not specify what the y-axis represents (percentage of models, count, or other); please state this explicitly in the caption or text.
  6. [References] Some references (e.g., Han et al. [24], Schneider et al. [25], Morrison et al. [26]) are given as bare arXiv numbers without full citation details; please standardize these entries to journal or arXiv format.

Circularity Check

1 steps flagged · score 4.0 of 10

One definitional step partially constructs the central transparency trend; no fitted parameter is renamed as a prediction.

  1. self definitional [Environmental Transparency Trends, category definitions and Figure 1 discussion]
    "Indirect Disclosure: Developers provided training compute data or released their model weights, allowing external estimates of training or inference impacts. ... From 2019 to 2022, transparency improved as awareness of impacts grew and open-weights model releases became more common. ... the introduction of increasingly commercial and proprietary models after 2022, potentially catalyzed by the popular launch of ChatGPT ... triggered a notable reversal in this trend, dramatically reducing direct environmental disclosures."

    The coding scheme defines 'Indirect Disclosure' as satisfied by releasing model weights even when no environmental data are provided, and the trend figure counts these categories as 'environmental impact transparency.' The paper's own account then explains the measured improvement and reversal by the timing of open-weights versus proprietary releases. The load-bearing conclusion that environmental transparency has declined since 2022 therefore reduces, by construction, to the classification of open-weight releases as transparency, rather than to an independently measured variable capturing disclosed environmental impact information. The OpenRouter usage percentages inherit this same category definition, though the no-disclosure share is also reported separately.

full rationale

The paper's central empirical contribution is a coding exercise over Epoch AI's Notable AI Models dataset plus OpenRouter traffic, not a fitted model; no parameter is estimated and then relabeled as a prediction. The main circularity risk is the self-definitional structure of the transparency categories: a model with open weights and no environmental data is placed in 'Indirect Disclosure,' so the time series largely re-expresses the known post-2022 shift from open research releases to closed commercial APIs. That reduces the headline 'declining transparency' trend in part by construction. However, substantive independent content remains: the OpenRouter 84% no-disclosure figure, the media spread analysis of the '3 Wh' estimate, and the tracing of the 'five cars' and '10% emissions' claims all stand on external data (OpenRouter, Google News, primary reports) rather than on the authors' fitted values. The authors do cite their own earlier measurement work and recommend the AI Energy Score, an author-maintained leaderboard, but those citations are used as methodological references and as a policy recommendation, not as an unverifiable load-bearing premise; they are a conflict-of-interest concern rather than circular reasoning. The definitional coding issue is thus a partial, real circularity in the trend claim, warranting a moderate score rather than a high one.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters or invented entities appear in this paper. The load-bearing premises are representativeness assumptions about the Epoch AI dataset, the OpenRouter sample, the Google News sample, the reliability of the manual coding, and the accuracy of the measurement tools recommended for disclosure.

assumptions (5)
  • domain assumption The Epoch AI Notable Models dataset is representative of the AI models that matter for assessing industry-wide transparency.
    The headline transparency trend is drawn solely from this curated dataset, which defines notable models as state of the art, highly cited, or historically notable; this selection may not reflect the broader population of deployed models (Section 'Environmental Transparency Trends').
  • domain assumption The three-category transparency classification is objective and applied consistently.
    The coding is performed by the authors without a published codebook or inter-rater reliability check, so classification decisions could vary and affect the trend (Section 'Environmental Transparency Trends').
  • domain assumption OpenRouter's top-20 monthly model usage is representative of overall LLM usage.
    The paper uses May 2025 OpenRouter data to claim that 84% of LLM usage flows through no-disclosure models, but OpenRouter is a single API platform and may not represent all LLM traffic (Figure 2).
  • domain assumption Google News search results for 'ChatGPT energy consumption' are representative of media coverage.
    The 100-article media analysis in 'Investigating the Urban Legends' rests on a convenience sample with no documented sampling strategy or stratification.
  • domain assumption Recommended measurement tools such as CodeCarbon and the AI Energy Score provide reliable enough data for disclosure.
    The recommendations promote these tools as the basis for reporting, but their accuracy, uncertainty, and comparability are not assessed in this paper (Section 'How to improve environmental impact disclosures in AI').

how reviews work

0 comments
Cite this review

Pith. "Pith review of Misinformation by Omission: The Need for More Environmental Transparency in AI." pith.science (2026). https://pith.science/paper/ZFWAIRKS

@misc{pith2026250615572,
  author       = {Pith},
  title        = {Pith review of: Misinformation by Omission: The Need for More Environmental Transparency in AI},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZFWAIRKS}},
  note         = {Machine review of arXiv:2506.15572}
}
read the original abstract

In recent years, Artificial Intelligence (AI) models have grown in size and complexity, driving greater demand for computational power and natural resources. In parallel to this trend, transparency around the costs and impacts of these models has decreased, meaning that the users of these technologies have little to no information about their resource demands and subsequent impacts on the environment. Despite this dearth of adequate data, escalating demand for figures quantifying AI's environmental impacts has led to numerous instances of misinformation evolving from inaccurate or de-contextualized best-effort estimates of greenhouse gas emissions. In this article, we explore pervasive myths and misconceptions shaping public understanding of AI's environmental impacts, tracing their origins and their spread in both the media and scientific publications. We discuss the importance of data transparency in clarifying misconceptions and mitigating these harms, and conclude with a set of recommendations for how AI developers and policymakers can leverage this information to mitigate negative impacts in the future.

Figures

Figures reproduced from arXiv: 2506.15572 by the authors.

Figure 1
Figure 1. Environmental Impact Transparency of Notable AI Models by Release Year27 yet common practice, although many articles accompanying papers did provide related information about, e.g. the amount of training data used or number of epochs trained. From 2019 to 2022, transparency improved as awareness of impacts grew and open-weights model releases became more common. This period includes the the work of Strubell et al.12… view at source ↗
Figure 2
Figure 2. Environmental Impact Transparency of LLM Usage – OpenRouter28 (May 2025) Beyond the long term trend, zooming in to examine recent AI model usage data helps illustrate today’s environmental impact transparency conditions. OpenRouter28, a widely-used API platform for LLMs, publicly shares data on model traffic including top 20 models by month, and the number of tokens running through every model. May 2025 data ( [PIT… view at source ↗
Figure 3
Figure 3. Analysis of media articles discussing ChatGPT energy consumption. AI can reduce 10% of global emissions While the numbers around AI’s negative environmental impacts can be misinterpreted and taken out of context, so, too, can the potential of AI to reduce emissions, especially by corporate actors that develop and deploy AI systems on a global scale. One recurring number states that AI can help reduce global GHG emis… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

37 extracted references · 23 canonical work pages

  1. [2]

    InProceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, 664–673 (2021)

    Kshirsagar, M.et al.Becoming good at AI for good. InProceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, 664–673 (2021). 3.DeepMind. Using ai to fight climate change (2009)

    work page 2021

  2. [4]

    & Kirchner, L

    Angwin, J., Larson, J., Mattu, S. & Kirchner, L. Machine bias. InEthics of data and analytics, 254–264 (Auerbach Publications, 2022)

    work page 2022

  3. [5]

    & Gebru, T

    Buolamwini, J. & Gebru, T. Gender shades: Intersectional accuracy disparities in commercial gender classification. In Conference on fairness, accountability and transparency, 77–91 (PMLR, 2018)

    work page 2018

  4. [6]

    InProceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency, 836–898 (2024)

    Rivera, J.-P.et al.Escalation risks from language models in military and diplomatic decision-making. InProceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency, 836–898 (2024). 7.Manheim, K. & Kaplan, L. Artificial intelligence: Risks to privacy and democracy.Yale JL & Tech.21, 106 (2019). 8.Summerfield, C.et al.How will advanc...

    arXiv 2024

  5. [9]

    Crawford, K.The atlas of AI: Power, politics, and the planetary costs of artificial intelligence(Yale University Press, 2021)

    work page 2021

  6. [10]

    M., Gebru, T., McMillan-Major, A

    Bender, E. M., Gebru, T., McMillan-Major, A. & Shmitchell, S. On the dangers of stochastic parrots: Can language models be too big? InProceedings of the 2021 ACM conference on fairness, accountability, and transparency, 610–623 (2021)

    2021

  7. [11]

    S., Viguier, S

    Luccioni, A. S., Viguier, S. & Ligozat, A.-L. Estimating the carbon footprint of BLOOM, a 176b parameter language model.arXiv preprint arXiv:2211.02001(2022)

    arXiv 2022

  8. [12]

    & McCallum, A

    Strubell, E., Ganesh, A. & McCallum, A. Energy and policy considerations for deep learning in NLP.arXiv preprint arXiv:1906.02243(2019)

    arXiv 2019

Show all 37 references
  1. [13]

    Luccioni, A. S. & Hernandez-Garcia, A. Counting carbon: A survey of factors influencing the emissions of machine learning.arXiv preprint arXiv:2302.08476(2023)

    2023 arXiv

  2. [14]

    InProceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, 1877–1894 (2022)

    Dodge, J.et al.Measuring the carbon intensity of AI in cloud instances. InProceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency, 1877–1894 (2022)

    2022

  3. [15]

    & Combaz, J

    Ligozat, A.-L., Lefèvre, J., Bugeau, A. & Combaz, J. Unraveling the hidden environmental impacts of AI solutions for environment.arXiv preprint arXiv:2110.11822(2021). 16.Patterson, D.et al.Carbon emissions and large neural network training.arXiv preprint arXiv:2104.10350(2021)

    2021 arXiv

  4. [17]

    In2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA), 854–867 (IEEE, 2021)

    Gupta, U.et al.Chasing Carbon: The Elusive Environmental Footprint of Computing. In2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA), 854–867 (IEEE, 2021)

    2021

  5. [18]

    K., Trivedi, H., Balasubramanian, A

    Cao, Q., Lal, Y . K., Trivedi, H., Balasubramanian, A. & Balasubramanian, N. IrEne: Interpretable energy prediction for transformers. In Zong, C., Xia, F., Li, W. & Navigli, R. (eds.)Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and th...

    doi

  6. [19]

    Wu, C.-J.et al.Sustainable AI: Environmental Implications, Challenges and Opportunities.arXiv preprint arXiv:2111.00364(2021)

    2021 arXiv

  7. [20]

    & Strubell, E

    Luccioni, S., Jernite, Y . & Strubell, E. Power hungry processing: Watts driving the cost of ai deployment? InThe 2024 ACM Conference on Fairness, Accountability, and Transparency, FAccT ’24, 85–99, DOI: 10.1145/3630106.3658542 (ACM, 2024)

    2024

  8. [21]

    & Gamazaychikov, B

    Luccioni, S. & Gamazaychikov, B. AI Energy Score Leaderboard. https://huggingface.co/spaces/AIEnergyScore/ Leaderboard (2025). Hugging Face Spaces. 22.Luccioni, S.et al.Light bulbs have energy ratings—so why can’t ai chatbots?Nature632, 736–738 (2024)

    2025

  9. [23]

    thirsty

    Li, P., Yang, J., Islam, M. A. & Ren, S. Making ai less" thirsty": Uncovering and addressing the secret water footprint of ai models.arXiv preprint arXiv:2304.03271(2023)

    2023 arXiv

  10. [24]

    & Ren, S

    Han, Y ., Wu, Z., Li, P., Wierman, A. & Ren, S. The unpaid toll: Quantifying the public health impact of ai.arXiv preprint arXiv:2412.06288(2024). 8/12

    2024 arXiv

  11. [25]

    2502.01671

    Schneider, I.et al.Life-cycle emissions of ai hardware: A cradle-to-grave approach and generational trends (2025). 2502.01671

    2025 arXiv

  12. [26]

    27.Epoch AI

    Morrison, J.et al.Holistically evaluating the environmental impact of creating language models.arXiv preprint arXiv:2503.05804(2025). 27.Epoch AI. Data on notable ai models (2024). Accessed: 2025-04-06

    2025 arXiv

  13. [28]

    Openrouter leaderboard rankings

    OpenRouter. Openrouter leaderboard rankings. https://openrouter.ai/rankings?view=month (2025). Accessed: 2025-06-03. 29.Lovins, A. B. Artificial intelligence meets natural stupidity: Managing the risks (2025). 30.Vaswani, A.et al.Attention is all you need.Adv. neural informati...

    2025

  14. [31]

    & Liang, C

    So, D., Le, Q. & Liang, C. The evolved transformer. In Chaudhuri, K. & Salakhutdinov, R. (eds.)Proceedings of the 36th International Conference on Machine Learning, vol. 97 ofProceedings of Machine Learning Research, 5877–5886 (PMLR, 2019)

    2019

  15. [32]

    Training a single ai model can emit as much carbon as five cars in their lifetimes.MIT technology Rev.75, 103 (2019)

    Hao, K. Training a single ai model can emit as much carbon as five cars in their lifetimes.MIT technology Rev.75, 103 (2019). 33.Toews, R. Deep learning’s carbon emissions problem.Forbes(2020). 34.Gemma Teamet al.Gemma 2: Improving open language models at a practical size (202...

    2019 arXiv

  16. [36]

    Kerr, D. & NPR. AI brings soaring emissions for Google and Microsoft, a major contributor to climate change.NPR, July (2024). 37.Chen, S. How much energy will ai really consume? the good, the bad and the unknown.Nature639, 22–24 (2025)

    2024

  17. [38]

    & Patel, P

    Aljbour, J., Wilson, T. & Patel, P. Powering intelligence: Analyzing artificial intelligence and data center energy consumption.EPRI White Pap. no. 3002028905(2024). 39.Dastin, J. & Nellis, S. Focus: For tech giants, ai like bing and bard poses billion-dollar search problem.Re...

    2024

  18. [42]

    ChatGPT, Deepseek & Co: How much energy do AI-powered chatbots consume? Euronews (online)

    Inês Trindade Pereira. ChatGPT, Deepseek & Co: How much energy do AI-powered chatbots consume? Euronews (online). Accessed: 2025-06-01

    2025

  19. [43]

    AI is bad for the environment, and the problem is bigger than energy consumption

    Hamish van der Ven. AI is bad for the environment, and the problem is bigger than energy consumption. The Conversation (online). Accessed: 2025-06-01

    2025

  20. [44]

    Data centers powering artificial intelligence could use more electricity than entire cities

    Spencer Kimball. Data centers powering artificial intelligence could use more electricity than entire cities. CNBC (online). Accessed: 2025-06-01

    2025

  21. [45]

    DeepSeek might not be such good news for energy after all

    James O’Donnell. DeepSeek might not be such good news for energy after all. MIT Technology Review (online). Accessed: 2025-06-01. 46.Berry Zwets. Researchers claim to cut energy consumption AI 95 percent. Techzine (online). Accessed: 2025-06-01. 47.Adam Clark Estes. Should you...

    2025

  22. [48]

    & Hutchinson, R

    Degot, C., Duranton, S., Frédeau, M. & Hutchinson, R. Reduce carbon and costs with the power of ai.Boston Consult. Group26(2021). 49.Dannouni, A.et al.Accelerating climate action with ai.Boston Consult. Group Special Rep. Google(2023)

    2021

  23. [50]

    Enabling the Net Zero transition: Assessing how the use of information and communication technology solutions impact greenhouse gas emissions of other sectors

    ITU. Enabling the Net Zero transition: Assessing how the use of information and communication technology solutions impact greenhouse gas emissions of other sectors. Tech. Rep. ITU-T L.1480, ITU (2022). Accessed: 2025-06-01. 51.WBCSD. Guidance on Avoided Emissions. Tech. Rep., ...

    2022

  24. [52]

    Das, K. P. & Chandra, J. A survey on artificial intelligence for reducing the climate footprint in healthcare.Energy Nexus 9, 100167 (2023). 53.The Environment. Artificial intelligence can reduce 5 to 10 percent ghg emission: Study (2022). 54.Kakkad, R. Google says AI could mi...

    2023

  25. [58]

    & Inouye, M

    Lannelongue, L., Grealey, J. & Inouye, M. Green algorithms: Quantifying the carbon footprint of computation.Adv. Sci. 2100707 (2021)

    2021

  26. [59]

    InProceedings of the conference on fairness, accountability, and transparency, 220–229 (2019)

    Mitchell, M.et al.Model cards for model reporting. InProceedings of the conference on fairness, accountability, and transparency, 220–229 (2019). 60.Luccioni, S. & Gamazaychikov, B. AI Models Hiding Their Energy Footprint? Here’s What You Can Do (2025)

    2019

  27. [61]

    environmental management380, 125008 (2025)

    Leal Filho, W.et al.European sustainability reporting standards: An assessment of requirements and preparedness of eu companies.J. environmental management380, 125008 (2025). 62.Gamazaychikov, B. Unveiling salesforce’s blueprint for sustainable ai: Where responsibility meets i...

    2025 arXiv

  28. [65]

    Llama 4 Model Card

    Meta. Llama 4 Model Card. https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct (2025). Hugging Face Model Card

    2025

  29. [66]

    Llama 3 Model Card

    Meta. Llama 3 Model Card. https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md (2024). GitHub Repository. Author contributions statement B.G. conducted the environmental impact transparency analysis, T.A.d.C carried out the media analysis. All authors wrote, edited and...

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.