Scenario-based forecasting of the global energy demand and carbon footprint of artificial intelligence

Berke M. Turkay; Ipek Pehlivan; Nuri C. Onat; Murat Kucukvar; Metin Türkay

doi:10.1371/journal.pone.0343056

Peer Review History

Original SubmissionJuly 14, 2025
19 Nov 2025 Decision Letter - John Adebisi, Editor PONE-D-25-38245Scenario-based forecasting of the global energy demand and carbon footprint of artificial intelligencePLOS ONE Dear Dr. Türkay, Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Reviewers have identified among other factors the need to add digital infrastructure sustainability, highlight the effect of scenario-based energy forecasting in carbon with respect to computation as well as some components of Materials and Methods. Details of of the reviewers report can be found below this email. Please submit your revised manuscript by Jan 03 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file. Please include the following items when submitting your revised manuscript: A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols. We look forward to receiving your revised manuscript. Kind regards, John Adebisi, Ph.D Academic Editor PLOS ONE Journal requirements: When submitting your revision, we need you to address these additional requirements. 1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf 2. We note that Figures 6, S1 Figure 6 and Figure 7 in your submission contain [map/satellite] images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright. We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission: 1. You may seek permission from the original copyright holder of Figures 6, S1 Figure 6 and Figure 7 to publish the content specifically under the CC BY 4.0 license. We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text: “I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.” Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission. In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].” 2. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only. The following resources for replacing copyrighted map figures may be helpful: USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/ The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/ Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/ Landsat: http://landsat.visibleearth.nasa.gov/ USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/# Natural Earth (public domain): http://www.naturalearthdata.com/ 3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise. Additional Editor Comments: Reviewers have recommended a major revision [Note: HTML markup is below. Please do not edit.] Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #1: Yes Reviewer #2: No ******** 2. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #2: No ****** 3. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #1: Yes Reviewer #2: Yes ****** 4. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #1: Yes Reviewer #2: Yes ****** 5. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #1: I recommend publication in PLOS ONE after major revision based on the following comments. 1. The reason for selection of digital infrastructure sustainability must be added. 2. The introduction is too verbose; special effort is needed to trip the contents without jeopardizing the message. 3.Author should highlights the effect of scenario-based energy forecasting in carbon intensity of computation. 4. The authors need to give more information about the research significance. 5. The role of artificial intelligence systems in carbon intensity of computation should be clarified. 6. Some important papers about sustainable energy (sush as doi.org/10.1039/D5EE02029H) can be cited. 7. In the conclusions part, only some conclusions were presented, beyond that, more perspectives are suggested to be provided. Reviewer #2: Title Unspecified, “Scenario-based forecasting of the global energy demand and carbon footprint of AI” fails to indicate the methodological novelty (e.g., MRIO integration, six-scenario modeling) or specific time horizon. Does not distinguish whether the analysis is empirical, simulation-based, or theoretical, reducing clarity for readers searching for methodological studies. Please add study design. Abstract Missing essential methodological transparency (e.g., model validation, uncertainty quantification). No clear mention of data sources or the calibration framework for efficiency parameters. Policy implications are asserted (“efficiency gains alone will not suffice”) but not supported by a quantitative comparison in the abstract. Sentence structure is long and complex, affecting readability. Keywords: Many and lacking prioritization. Some are descriptive rather than functional (“technology adoption” is unclear). Missing terms that reflect the study’s modeling approach (e.g., “scenario analysis,” “sustainability modeling,” “AI energy systems”). Introduction Lengthy and redundant. The background has many pages and repeats data center energy statistics. Limited critical synthesis of prior scenario-based energy studies or integrated assessment models (IAMs); references 5-8 are not analyzed comparatively. The knowledge gap is not articulated; it is unclear whether the novelty lies in global scope, scenario design, or MRIO coupling. Unsupported generalizations, such as “AI currently represents approximately 0.4% of global electricity consumption,” are presented without citing precise datasets or year references. No clear conceptual framework figure to orient the reader before reading the “Methods.” Materials and Methods Model Framework The structure (Fig. 1) is descriptive but there are no validation steps or uncertainty propagation techniques (e.g., Monte Carlo or sensitivity analysis). No justification for selecting exponential and logistic functional forms for model scaling and efficiency gains. Energy Efficiency Modeling Equations (1) to (4) are complex but empirical parameters are not justified nor referenced. Independence assumptions among hardware, algorithmic, and infrastructural efficiencies (Eq. 4) are strong and unsupported. No validation using real-world benchmarks (e.g., GPU generation trends or data center efficiency reports). Computational Demand Depends on compound exponential extrapolation of model complexity and query volume. Population and adoption rates (Eq. 10) use logistic diffusion but no region-specific calibration is shown. Data sources unclear (Supplementary Sections cited but not summarized in text). MRIO Analysis The MRIO methodology is standard, but no sensitivity testing or temporal updating of EXIOBASE is described. Attribution of GPU manufacturing entirely to Taiwan is over-simplified; ignores TSMC Arizona, Samsung Texas, or emerging EU fabs. Carbon Footprint Estimation Equations (15) to (21) are mathematically correct but no validation against known industry LCA benchmarks (e.g., ISO 14040 standards). No error propagation or confidence intervals reported for emission totals. Results Global Energy Demand Scenarios Results are presented as deterministic forecasts though input assumptions are random. Carbon Footprint Breakdown No benchmarking against existing sectoral projections (IEA, IPCC, or McKinsey AI energy reports). Country-level disparities are interesting but derived from proxy variables (IHDI) instead of actual digital infrastructure metrics. Tables referenced in Supplementary Information are essential to credibility but not summarized in the main text. Tables and Figures Tables are descriptive but do not include units, uncertainty, or data sources. Figures 2-3 show projections but lack axis units and confidence interval methods. Figures 2-6 lack legends detailing whether shaded bands represent model variance or scenario spread. Results in Figs 4-6 depend on unvalidated parameters. Discussion Reads more like general ideas instead of being supported by data, analysis, or evidence. Insufficient linkage to numerical findings. Circular reasoning: claims that “efficiency gains alone will not suffice” rest on model assumptions, not strong validation. Exclusion of economic and policy feasibility dimensions weakens the practical relevance. Several citations (30-35) are referenced but not elaborated; end-of-life and circular economy aspects are superficial. Conclusion Mostly a repetition of limitations. No actionable policy guidance or quantifiable thresholds (e.g., emission reduction targets, scenario prioritization). Shifts from cautious projection to assertive prediction (“AI workloads could consume 30 % of global electricity”) without strong evidence. References Formatting inconsistent. Depends on secondary sources (Forbes, Statista-type data) rather than peer-reviewed energy or computing journals. Insufficient citation of existing MRIO and integrated assessment models (e.g., GTAP-E, GCAM, or IAMC frameworks). ****** 6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #1: No Reviewer #2: No ******** [NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.] To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation. NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications. Attachments Attachment Submitted filename: reviewer comments.docx https://doi.org/10.1371/journal.pone.0343056.r001
Revision 1
20 Jan 2026 Author Response We thank the reviewers for their careful reading of the manuscript and for their constructive comments. We have made the appropriate changes to formatting, including using world maps in the public domain. Below, we respond to each of the reviewer’s comments point by point. Reviewer #1: I recommend publication in PLOS ONE after major revision based on the following comments. 1. The reason for the selection of digital infrastructure sustainability must be added. The revision clarifies that digital infrastructure, specifically data centers, is the focal point because it is the primary source of energy consumption and carbon emissions associated with AI systems (please see page 1, the second paragraph of the introduction). 2. The introduction is too verbose; special effort is needed to trip the contents without jeopardizing the message. The introduction has been restructured to reduce redundancy and improve focus, while preserving the necessary contextual grounding for a global, long-horizon systems analysis. 3. The authors should highlight the effect of scenario-based energy forecasting in carbon intensity of computation. This effect is now explicitly emphasized. The revised manuscript demonstrates quantitatively that deployment structure and electricity mix assumptions can dominate efficiency gains in determining the carbon intensity of computation, a central result illustrated in the scenario comparisons (please see page 3, the final paragraph of the introduction). 4. The authors need to give more information about the research significance. The revised introduction explicitly identifies the knowledge gap addressed by this study: the absence of an integrated framework that jointly models global AI deployment trajectories, evolving electricity systems, and supply-chain emissions. The novelty lies in coupling scenario-based computational demand modeling with MRIO-based life-cycle carbon accounting, rather than in global scope alone. 5. The role of artificial intelligence systems in the carbon intensity of computation should be clarified. The revised text clarifies that AI systems influence carbon intensity by determining where, when, and at what scale electricity is consumed, through choices in model architecture, deployment patterns, and reliance on centralized versus distributed infrastructure (please see page 2, the fourth paragraph of the introduction). 6. Some important papers about sustainable energy (such as doi.org/10.1039/D5EE02029H) can be cited. Relevant sustainable energy systems literature, including the suggested article, has been incorporated into the revised manuscript. The new citations include: • Sanati et al. (2025), which informs the discussion of energy system efficiency improvements and low-carbon innovation pathways relevant to long-run decarbonization scenarios. • Yang et al. (2025), which supports the treatment of embodied renewable energy and global trade linkages within the multi-regional input–output framework. • Lu and Chen (2025), which provides methodological grounding for the spatially disaggregated MRIO modeling of power sector emissions used in the regional analysis. • Faturay et al. (2020), which contextualizes the application of MRIO models to energy system transitions and supports the treatment of upstream economic and energy impacts. 7. In the conclusions part, only some conclusions were presented, beyond that, more perspectives are suggested to be provided. The Conclusion has been revised to more clearly synthesize the study’s core findings and their implications, rather than focusing primarily on methodological limitations. The updated text highlights that AI’s long-term energy use and emissions are governed by the interaction between deployment structure, electricity system decarbonization, and the geographic concentration of compute, with efficiency improvements alone proving insufficient. It now summarizes key contrasts across scenarios, emphasizing the comparatively larger emissions reductions achievable through structural interventions such as model consolidation and low-carbon electricity transitions. Remaining uncertainties and data limitations inherent to long-horizon scenario analysis are retained but more directly linked to the interpretation of results, and the exploratory (non- predictive) nature of the scenarios is clarified alongside the resulting policy relevance. Reviewer #2: Title Unspecified, “Scenario-based forecasting of the global energy demand and carbon footprint of AI” fails to indicate the methodological novelty (e.g., MRIO integration, six- scenario modeling) or specific time horizon. Does not distinguish whether the analysis is empirical, simulation-based, or theoretical, reducing clarity for readers searching for methodological studies. Please add study design. The revised abstract explicitly characterizes the study as a “scenario-based, simulation-driven modeling framework” that links mathematical representations of computational demand with life-cycle carbon accounting. This clearly distinguishes the work from purely empirical or theoretical studies. Abstract Missing essential methodological transparency (e.g., model validation, uncertainty quantification). No clear mention of data sources or the calibration framework for efficiency parameters. Policy implications are asserted (“efficiency gains alone will not suffice”) but not supported by a quantitative comparison in the abstract. Sentence structure is long and complex, affecting readability. Keywords: Many and lacking prioritization. Some are descriptive rather than functional (“technology adoption” is unclear). Missing terms that reflect the study’s modeling approach (e.g., “scenario analysis,” “sustainability modeling,” “AI energy systems”). The revised abstract now includes explicit quantitative comparisons. In particular, it reports that deployment and electricity-mix scenarios can reduce total emissions by up to 40% relative to business-as-usual pathways, exceeding the reductions achievable through efficiency gains alone by more than 20 percentage points. These figures directly support the stated policy implications. The keyword list has been revised to prioritize methodological and analytical content. Terms such as scenario analysis, electricity demand forecasting, AI energy systems, and multi-regional input-output analysis have been added, while overly generic descriptors have been removed. Introduction Lengthy and redundant. The background has many pages and repeats data center energy statistics. The introduction has been restructured to reduce redundancy and improve focus, while preserving the necessary contextual grounding for a global, long-horizon systems analysis. Limited critical synthesis of prior scenario-based energy studies or integrated assessment models (IAMs); references 5-8 are not analyzed comparatively. The cited references were not intended to represent a set of comparable scenario- based or IAM studies. Rather than synthesizing existing IAM projections, our objective is to introduce a complementary, infrastructure-focused modeling framework that links AI deployment to data-center electricity demand and upstream supply chains. The cited works therefore serve to motivate the problem space and highlight gaps in existing assessments, which the proposed MRIO-based approach is designed to address. The knowledge gap is not articulated; it is unclear whether the novelty lies in global scope, scenario design, or MRIO coupling. The revised introduction explicitly identifies the knowledge gap addressed by this study: the absence of an integrated framework that jointly models global AI deployment trajectories, evolving electricity systems, and supply-chain emissions. The novelty lies in coupling scenario-based computational demand modeling with MRIO-based life-cycle carbon accounting, rather than in global scope alone. Unsupported generalizations, such as “AI currently represents approximately 0.4% of global electricity consumption,” are presented without citing precise datasets or year references. This estimate is now explicitly anchored to the year 2024 and supported by Shehabi, Smith, and Masanet’s paper published by the Lawrence Berkeley National Laboratory (see page 1, second paragraph of the introduction). The temporal and empirical basis of the figure is clearly stated. No clear conceptual framework figure to orient the reader before reading the “Methods.” A detailed conceptual and methodological framework is already provided as a comprehensive figure at the beginning of the Methods section, where it can be interpreted alongside the formal model definitions. We intentionally avoided introducing an additional high-level schematic earlier in the manuscript to prevent redundancy and ensure that the framework is presented with sufficient technical context. Materials and Methods Model Framework The structure (Fig. 1) is descriptive but there are no validation steps or uncertainty propagation techniques (e.g., Monte Carlo or sensitivity analysis). Figure 1 is intended to present the conceptual and modular structure of the modeling framework rather than to serve as a validation or uncertainty analysis figure. Validation in the conventional predictive sense is not feasible in a long-horizon, scenario-based assessment, as the model is not designed to forecast realizations but to bound plausible futures conditional on structural assumptions. Uncertainty is instead incorporated through explicit scenario design and parametric ranges calibrated from the literature, with resulting dispersion reported in downstream figures. We have clarified this distinction in the revised manuscript and explicitly state that the framework is exploratory and comparative rather than predictive. A discussion of uncertainty sources and their propagation through scenario spread has also been expanded in the Methods and Discussion sections. No justification for selecting exponential and logistic functional forms for model scaling and efficiency gains. The justification for the use of exponential and logistic functional forms has been made explicit in the Methods section, primarily in Sections “Energy efficiency” and “Computational demand.” These sections now explain that decelerating exponential forms are used to represent empirically observed patterns of rapid early improvement followed by diminishing returns in hardware, algorithmic, and infrastructure efficiency, while logistic diffusion is employed to capture saturation effects in AI adoption. The text clarifies that these functions are used as stylized bounds for long-run scenario analysis rather than as short-term predictive models. Energy Efficiency Modeling Equations (1) to (4) are complex but empirical parameters are not justified nor referenced. All empirical parameters are now explicitly documented and referenced in the Methods section at first use, with detailed justification provided in the Supplementary Sections 2.1-2.3 (efficiency parameters) and 3.1–3.4 (computational demand parameters). These supplementary sections describe data sources, calibration procedures, parameter ranges, and sensitivity bounds, and are cross-referenced in the main text to ensure transparency and reproducibility. Independence assumptions among hardware, algorithmic, and infrastructural efficiencies (Eq. 4) are strong and unsupported. Independence among hardware, algorithmic, and infrastructural efficiency components is a standard first-order modeling assumption adopted for transparency and tractability. The limitation of this assumption is explicitly acknowledged, and relaxing it would require data that are currently unavailable at a global scale. No validation using real-world benchmarks (e.g., GPU generation trends or data center efficiency reports). The model is calibrated against historical GPU efficiency trends and reported data center PUE improvements. Long-horizon validation against future outcomes is not feasible for scenario-based projections through 2050. The purpose of the framework is comparative scenario analysis rather than predictive validation. Computational Demand Depends on compound exponential extrapolation of model complexity and query volume. The use of multiple contrasting scenarios is precisely intended to bound realistic growth trajectories rather than assert a single extrapolation. Scenario-based modeling is employed to explore structural sensitivities, not to claim deterministic forecasts. Population and adoption rates (Eq. 10) use logistic diffusion but no region-specific calibration is shown. Regional heterogeneity is incorporated through stratification by inequality-adjusted human development index (IHDI), which serves as a transparent and globally available proxy (see Supplementary Section 3.3.1). Country-specific digital infrastructure calibration is incorporated by accounting for the number of data centers, as outlined in the Supplementary Section 7.2.1. Data sources unclear (Supplementary Sections cited but not summarized in text). MRIO Analysis Response: We thank the reviewer for this comment and agree that clarity regarding data sources is important for transparency and reproducibility. In the revised manuscript, we have added a concise summary statement to the Multi-Regional Input-Output Model section explicitly outlining the key data inputs used in the MRIO framework (including electricity production values, emission factors, price tables, and country-specific adjustments), while retaining detailed numerical tables and sectoral mappings in the Supplementary Information. This ensures that readers can readily identify the provenance and structure of the data directly from the main text without consulting supplementary materials. Action: A one-sentence summary of MRIO data sources has been added to Section 7.1.2. The MRIO methodology is standard, but no sensitivity testing or temporal updating of EXIOBASE is described. Response: In this study, the MRIO framework is not used as a forecasting or scenario-generating model, but rather as a carbon accounting layer that translates scenario-driven energy demand trajectories into associated life-cycle emissions. Uncertainty and sensitivity in this study are addressed primarily through the design of the scenario framework rather than through sensitivity analysis at the MRIO calculations. The scenarios systematically vary the key drivers of AI-related energy demand and emissions, including regional AI adoption rates and diffusion pathways, alternative model scaling configurations (Baseline, Fewer–Larger, and More–Smaller), projected trajectories of hardware and algorithmic efficiency improvements, and contrasting electricity system assumptions under Business-as-Usual and Energy Target pathways. By exploring these dimensions in combination, the scenario framework captures the dominant sources of uncertainty influencing long-term AI energy demand and associated carbon emissions in a transparent and policy-relevant manner. These dimensions constitute the primary analytical focus of the paper. Conducting additional MRIO-level sensitivity analysis would therefore be methodologically redundant and risk conflating accounting uncertainty with scenario uncertainty, which the framework deliberately separates. Regarding temporal updating, the analysis employs EXIOBASE 3.8.2, the most recent publicly available release, reflecting contemporary global production technologies and trade structures. In this study, MRIO coefficients are applied as time-invariant emission intensities to scenario-specific energy demand, rather than being endogenously projected over time. As such, temporal interpolation or extrapolation of MRIO tables would not meaningfully improve the robustness of results and would introduce additional, largely unverifiable assumptions about future global supply-chain restructuring. Moreover, for key sectors driving results, most notably electricity generation, the emission intensities used in EXIOBASE are grounded in physical and thermodynamic characteristics (e.g., fuel carbon content and conversion efficiencies). These electricity generation sectors’ direct emissions are most Attachments Attachment Submitted filename: Response_to_Reviewers.pdf https://doi.org/10.1371/journal.pone.0343056.r002
1 Feb 2026 Decision Letter - John Adebisi, Editor Scenario-based forecasting of the global energy demand and carbon footprint of artificial intelligence PONE-D-25-38245R1 Dear Dr. Türkay, We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements. Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication. However It has been observed that there is a change in authorship, which creates disparity between the initial authors during submission, review and now. please read the plus one conditions regarding this and reach out to the the journal for clarifications. An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support. If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. Kind regards, John Adebisi, Ph.D Academic Editor PLOS One Additional Editor Comments (optional): Reviewers' comments: Reviewer's Responses to Questions Comments to the Author 1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation. Reviewer #1: (No Response) Reviewer #2: All comments have been addressed ******** 2. Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. Reviewer #1: Yes Reviewer #2: Yes ****** 3. Has the statistical analysis been performed appropriately and rigorously? Reviewer #1: Yes Reviewer #2: Yes ****** 4. Have the authors made all data underlying the findings in their manuscript fully available? The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified. Reviewer #1: Yes Reviewer #2: Yes ****** 5. Is the manuscript presented in an intelligible fashion and written in standard English? PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Reviewer #1: Yes Reviewer #2: Yes ****** 6. Review Comments to the Author Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters) Reviewer #1: The authors answered the comments very well. Therefore, I recommend this revised manuscript for publication in its current form. Reviewer #2: I am satisfied that the authors have carefully and comprehensively addressed all major and minor comments raised in the previous review round. ****** 7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files. If you choose “no”, your identity will remain anonymous but your review may still be made public. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Reviewer #1: No Reviewer #2: No ******** https://doi.org/10.1371/journal.pone.0343056.r003
Formally Accepted
Acceptance Letter - John Adebisi, Editor PONE-D-25-38245R1 PLOS One Dear Dr. Türkay, I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team. At this stage, our production department will prepare your paper for publication. This includes ensuring the following: * All references, tables, and figures are properly cited * All relevant supporting information is included in the manuscript submission, * There are no issues that prevent the paper from being properly typeset You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps. Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org. You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing. If we can help with anything else, please email us at customercare@plos.org. Thank you for submitting your work to PLOS ONE and supporting open access. Kind regards, PLOS ONE Editorial Office Staff on behalf of Dr. John Adebisi Academic Editor PLOS One https://doi.org/10.1371/journal.pone.0343056.r004

Open letter on the publication of peer review reports

PLOS recognizes the benefits of transparency in the peer review process. Therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. Reviewers remain anonymous, unless they choose to reveal their names.

We encourage other journals to join us in this initiative. We hope that our action inspires the community, including researchers, research funders, and research institutions, to recognize the benefits of published peer review reports for all parts of the research system.

Learn more at ASAPbio .