IA verda i la Llei IA: Legislació pionera o simple declaració d’intencions en matèria mediambiental?
Número
Secció
Paraules clau:
Green AI , Acte IA , Intel•ligència Artificial , Unió Europea , Sostenibilitat , Centres de dades
Publicat
Resum
Aquest article examina com ha evolucionat la perspectiva mediambiental de la Unió Europea sobre la intel·ligència artificial, tot contrastant les elevades expectatives fixades pel Pacte Verd Europeu amb l’Acte IA que el va seguir. Malgrat que el Pacte Verd reconeixia el potencial de la IA per millorar l’eficiència energètica, no va abordar de manera exhaustiva el consum d’aigua i d’energia ni la gestió dels residus electrònics derivats del desenvolupament de grans models d’aprenentatge profund. Posteriorment, el Llibre Blanc sobre IA va aprofundir més en la dimensió ambiental d’aquesta tecnologia; tanmateix, l’Acte IA —aprovat el 2024— no tradueix completament aquestes propostes en obligacions concretes. L’article analitza les principals disposicions de l’Acte IA relacionades amb la sostenibilitat, incidint en l’absència de mecanismes directes per limitar el consum energètic, mitigar la petjada hídrica o garantir una gestió adequada dels residus electrònics. En aquest context, es proposen dues mesures de lege ferenda per resoldre aquestes mancances: la inclusió obligatòria de factors d’impacte ambiental en els benchmarks competitius de IA, i la implementació d’un sistema d’etiquetatge mediambiental que informi els consumidors sobre la sostenibilitat dels centres de dades on operen els models.
Descàrregues
Agències de suport
NoLlicència
Drets d'autor (c) 2025 Revista Catalana de Dret Ambiental
Aquesta obra està sota una llicència internacional Creative Commons Reconeixement-NoComercial-SenseObraDerivada 4.0.
Els autors que publiquen en aquesta revista estan d’acord amb els termes següents:- Els autors conserven els drets d’autoria i atorguen a la revista el dret de primera publicació, amb l’obra disponible simultàniament sota una Llicència d’Atribució, No comercial i sense derivades CC BY NC ND de CreativeCommons que permet compartir l’obra amb tercers, sempre que aquests en reconeguin l’autoria i la publicació inicial en aquesta revista, utilitzar materials sense fins comercials i crear materials sense distribuir el material modificat.
- Els autors són lliures de fer acords contractuals addicionals independents per a la distribució no exclusiva de la versió de l’obra publicada a la revista (com ara la publicació en un repositori institucional o en un llibre), sempre que se’n reconegui la publicació inicial en aquesta revista.
- S’encoratja els autors a publicar la seva obra en línia (en repositoris institucionals o a la seva pàgina web, per exemple) abans i durant el procés de tramesa, amb l’objectiu d’aconseguir intercanvis productius i fer que l’obra obtingui més citacions (vegeu The Effect of Open Access, en anglès).
Referències
Alder, N., Ebert, K., Herbrich, R., & Hacker, P. (2024). AI, climate, and transparency: Operationalizing and improving the AI act. arXiv preprint arXiv:2409.07471. https://doi.org/10.48550/arXiv.2409.07471
Almeida, D. V., Kolinjivadi, V., Ferrando, T., & Roy, B. (2023). The “greening” of empire: The European Green Deal as the EU first agenda. Political Geography, 96. https://doi.org/10.1016/j.polgeo.2023.102925
Apráez, B. E., & Noorman, M. (2024). Regulating AI in the ‘twin transitions’: Significance and shortcomings of the AI Act in the digitalised electricity sector. Review of European, Comparative & International Environmental Law, 367. https://doi.org/10.1111/reel.12574
Aşnaz, M. S. K. (2024, June). AI-driven energy management: Innovative solutions for achieving green deal targets in industry. In 5th International Eurasian Conference on Science, Engineering, and Technology (EurasianSciEnTech 2024), p. 685. Retrieved from <https://www.researchgate.net/profile/Melike-Karasu-Asnaz/publication/382306269_AI-driven_energy_management_innovative_solutions_for_achieving_green_deal_targets_in_industry/links/66979e304a172d2988a6368e/AI-driven-energy-management-innovative-solutions-for-achieving-green-deal-targets-in-industry.pdf> [Last consulted: January 18, 2025]
Azarifar, M., Arik, M., & Chang, J. Y. (2024). Liquid cooling of data centers: A necessity facing challenges. Applied Thermal Engineering, 123112. http://doi.org/10.1016/j.applthermaleng.2024.123112
Beloglazov, A., Buyya, R., Lee, Y. C., & Zomaya, A. (2011). A taxonomy and survey of energy-efficient data centers and cloud computing systems. Advances in Computers, 82, pp. 47–111. https://doi.org/10.1016/B978-0-12-385512-1.00003-7
Bolón-Canedo, V., Morán-Fernández, L., Cancela, B., & Alonso-Betanzos, A. (2024). A review of green artificial intelligence: Towards a more sustainable future. Neurocomputing, 128096, p. 7. https://doi.org/10.1016/j.neucom.2024.128096
Bowman, S. R., Dahl, G. E., De Marneffe, M.-C., Frost, R., Jia, R., Livingston, K., Potts, C., & Rasooli, M. S. (2021). What will it take to fix benchmarking in natural language understanding? arXiv preprint arXiv:2104.02145, p. 6. https://doi.org/10.48550/arXiv.2104.02145
Brownsword, R. (2019). Law, technology and society: Reimagining the regulatory environment. Routledge. Retrieved from <https://www.routledge.com/Law-Technology-and-Society-Reimagining-the-Regulatory-Environment/Brownsword/p/book/9780815356462>
Bucilua, C., Caruana, R., & Niculescu-Mizil, A. (2006). Model compression. In KDD '06: Proceedings of the 12th ACM SIGKDD international conference on knowledge discovery and data mining, pp. 535–541. https://doi.org/10.1145/1150402.1150464
Cao, Z., Zhou, X., Hu, H., Wang, Z., & Wen, Y. (2022). Toward a systematic survey for carbon neutral data centers. IEEE Communications Surveys & Tutorials, 24(2), pp. 895–936. https://doi.org/10.48550/arXiv.2110.09284
Campins Eritja, M. (2018, May). Competències i principis del Dret Ambiental de la UE. [Conference presentation]. Seminari: Formació en Dret Ambiental de la UE, Càtedra Jean Monnet de Dret Ambiental de la UE, Escola Judicial del CGPJ de Barcelona.
Coeckelbergh, M. (2021). AI for climate: Freedom, justice, and other ethical and political challenges. AI and Ethics, 1(1), pp. 67–72. https://doi.org/10.1007/s43681-020-00007-2
Corrigan, C., & Lucaj, L. (2020). The potential for AI in implementing the Green Deal and ethical implications. Institute for Ethics in Artificial Intelligence. Retrieved from <https://ieai.sot.tum.de/wp-content/uploads/2022/03/Research-Brief_GreenDeal_Final_Update.pdf>
Daneshjou, R., Smith, M. P., Sun, M. D., Rotemberg, V., & Zou, J. (2021). Lack of transparency and potential bias in artificial intelligence data sets and algorithms: A scoping review. JAMA Dermatology, 157(11), pp. 1362–1369. https://doi.org/10.1001/jamadermatol.2021.3129
Dhar, P. (2020). The carbon impact of artificial intelligence. Nature Machine Intelligence, 2(8), pp. 423–425. https://doi.org/10.1038/s42256-020-0219-9
Ebert, K., Alder, N., Herbrich, R., & Hacker, P. (2024). AI, climate, and regulation: From data centers to the AI Act. arXiv preprint arXiv:2410.06681, p. 4. https://doi.org/10.48550/arXiv.2410.06681
European Commission. (2020). Communication: Shaping Europe’s digital future. Retrieved from <https://commission.europa.eu/system/files/2020-02/communication-shaping-europes-digital-future-feb2020_en_4.pdf>
European Commission. (2021). Investing to shape our future. Research and Innovation. Retrieved from <https://research-and-innovation.ec.europa.eu/document/download/9224c3b4-f529-4b48-b21b-879c442002a2_en?filename=ec_rtd_he-investing-to-shape-our-future.pdf>
European Commission. (2024, August 8). Artificial Intelligence Act: Call for tenders to measure and foster energy efficient and low emission artificial intelligence in the EU [Call for tenders]. Digital Strategy. Retrieved from <https://digital-strategy.ec.europa.eu/en/funding/artificial-intelligence-act-call-tenders-measure-and-foster-energy-efficient-and-low-emission>
European Parliament. (2020, February 19). Carbon footprint of artificial intelligence (AI): Answer to written question E-001000/2020 to the Commission (Rule 138). Retrieved from <https://www.europarl.europa.eu/doceo/document/E-9-2020-001000_EN.html>
European Parliament. (2024, March 12). Answer to written question E-003690/2023(ASW) by Mr. Breton on behalf of the European Commission [Written question]. Retrieved from <https://www.europarl.europa.eu/doceo/document/E-9-2023-003690-ASW_EN.html>
European Parliament. (2024, October 7). Artificial intelligence (AI) and energy consumption: Priority question for written answer P-001974/2024 to the Commission (Rule 144). Retrieved from <https://www.europarl.europa.eu/doceo/document/P-10-2024-001974_EN.html>
European Parliament. (2024, October 8). AI and energy consumption: Answer in writing – Question for written answer E-001977/2024 to the Commission (Rule 144). Retrieved from <https://www.europarl.europa.eu/doceo/document/E-10-2024-001977_EN.html>
European Parliament. (2024, November 25). AI and the implementation of the Digital Markets Act: Priority question for written answer P-002649/2024 to the Commission (Rule 144). Retrieved from <https://www.europarl.europa.eu/doceo/document/P-10-2024-002649_EN.html>
European Parliament. (2024, December 2). Openness in AI and transparency of training data: Question for written answer E-002709/2024 to the Commission (Rule 144). Retrieved from <https://www.europarl.europa.eu/doceo/document/E-10-2024-002709_EN.html>
Frankle, J., Dziugaite, G. K., Roy, D. M., & Carbin, M. (2020). Pruning neural networks at initialization: Why are we missing the mark?. arXiv preprint arXiv:2009.08576. https://doi.org/10.48550/arXiv.1611.06440
Frankle, J., & Carbin, M. (2019). The lottery ticket hypothesis: Finding sparse, trainable neural networks. In International Conference on Learning Representations (ICLR). arXiv preprint arXiv:1803.03635. https://doi.org/10.48550/arXiv.1803.03635
Gailhofer, P., Herold, A., Schemmel, J. P., Scherf, C. S., de Stebelski, C. U., Köhler, A. R., & Braungardt, S. (2021). The role of artificial intelligence in the European green deal. European Parliament, p. 30. https://doi.org/10.1109/TEM.2024.3469370
Haigh, N. (2015). EU environmental policy: Its journey to centre stage,p. 31. Routledge. https://doi.org/10.4324/9781315712475
Hacker, P. (2024). Sustainable AI regulation. Common Market Law Review, 61(2), 2. https://doi.org/10.54648/cola2024025
Hainsch, K., Löffler, K., Burandt, T., & Auer, H. (2022). Energy transition scenarios: What policies, societal attitudes, and technology developments will realize the EU Green Deal? Energy, 239. https://doi.org/10.1016/j.energy.2021.122067
Han, S., Mao, H., & Dally, W. J. (2016). Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding. 4th International Conference on Learning Representations (ICLR). arXiv preprint arXiv:1510.00149
Han, S., Pool, J., Tran, J., & Dally, W. (2015). Learning both weights and connections for efficient neural network. Advances in Neural Information Processing Systems, 28. https://dl.acm.org/doi/10.5555/2969239.2969366
Harlow, C., & Rawlings, R. (2009). Law and administration (3rd ed., p. 195). Cambridge University Press. Retrieved from <https://ssrn.com/abstract=1490038>
Henderson, P., Hu, J., Romoff, J., Brunskill, E., Jurafsky, D., & Pineau, J. (2020). Towards the systematic reporting of the energy and carbon footprints of machine learning. Journal of Machine Learning Research, 21(248), pp. 1–43. (Preliminary version: arXiv:2002.05651)
Heymann, F., Parginos, K., Bessa, R. J., & Galus, M. (2023). Operating AI systems in the electricity sector under European’s AI Act–Insights on compliance costs, profitability frontiers and extraterritorial effects. Energy, 10, pp. 4538–4555. https://doi.org/10.1016/j.egyr.2023.11.020
Hinton, G., Vinyals, O., & Dean, J. (2015). Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531, p. 3. Retrieved from <https://arxiv.org/abs/1503.02531>
Jordan, A. J., & Lenschow, A. (2010). Environmental policy integration: A state of the art review. Environmental Policy and Governance, 20(3), pp. 147–158. https://doi.org/10.1002/eet.539
Jin, X., Zhang, F., Vasilakos, A. V., & Liu, Z. (2016). Green data centers: A survey, perspectives, and future directions. arXiv preprint arXiv:1608.00687. https://doi.org/10.48550/arXiv.1608.00687
Kelemen, R. D., & Pavone, T. (2022). Where have the guardians gone? Law enforcement and the politics of supranational forbearance in the European Union. American Political Science Review, 116(1), pp. 291–307. http://dx.doi.org/10.2139/ssrn.3994918
Kiddee, P., Naidu, R., & Wong, M. H. (2013). Electronic waste management approaches: An overview. Waste Management, 33(5), pp. 1237–1250. https://doi.org/10.1016/j.wasman.2013.01.006
Kougias, I., Taylor, N., & Kakoulaki, G. (2021). The role of photovoltaics for the European Green Deal and the recovery plan. Renewable and Sustainable Energy Reviews, 150. https://doi.org/10.1016/j.rser.2021.111017
Koundouri, P., Devves, S., & Plataniotis, A. (2021). Alignment of the European green deal, the sustainable development goals and the European semester process: Method and application. Theoretical Economics Letters, 11(4), pp. 743–770. https://doi.org/10.4236/tel.2021.114049
Leonard, M., Pisani-Ferry, J., Shapiro, J., Tagliapietra, S., & Wolff, G. B. (2021). The geopolitics of the European green deal (No. 04/2021) (p. 12). Bruegel. Retrieved from <https://ecfr.eu/wp-content/uploads/The-geopolitics-of-the-European-Green-Deal.pdf>
Li, P., Yang, J., Islam, M. A., & Ren, S. (2023). Making AI less “thirsty”: Uncovering and addressing the secret water footprint of AI models. arXiv preprint arXiv:2304.03271. https://doi.org/10.48550/arXiv.2304.03271
Lilkov, D. (2021). Regulating artificial intelligence in the EU: A risky game. European View, 20(2), pp. 166–174. https://doi.org/10.1177/17816858211059248
Liu, H. M., & Chang, Z. (2024). Multi-objective optimization of temperature uniformity in the immersion liquid cooling cabinet with Taguchi-based grey relational analysis. International Communications in Heat and Mass Transfer, 154, 107395. https://doi.org/10.1016/j.icheatmasstransfer.2024.107395
Liu, X., Lu, D., Zhang, A., Liu, Q., & Jiang, G. (2022). Data-driven machine learning in environmental pollution: Gains and problems. Environmental Science & Technology, 56(4), pp. 2124–2133. https://doi.org/10.1021/acs.est.1c06157
Majone, G. (2014). From regulatory state to a democratic default. JCMS: Journal of Common Market Studies, 52(6), pp. 1216–1223. https://doi.org/10.1111/jcms.12190
Molchanov, P., Tyree, S., Karras, T., Aila, T., & Kautz, J. (2017). Pruning convolutional neural networks for resource efficient inference. 5th International Conference on Learning Representations (ICLR). arXiv preprint arXiv:1611.06440. https://doi.org/10.48550/arXiv.1611.06440
Nassar, D. (2025). A holistic approach to addressing environmental sustainability in data centers (Doctoral dissertation, University of East London). 26. https://doi.org/10.15123/uel.8y6q9
Ottman, J. (2017). The new rules of green marketing: Strategies, tools, and inspiration for sustainable branding (p. 22). Routledge. https://doi.org/10.4324/9781351278683
European Parliament and Council of the European Union (2024). Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonized rules in the field of artificial intelligence.... (Official Journal of the European Union).
Patterson, D., Gonzalez, J., Le, Q., Liang, C., Munguia, L. M., Rothchild, D., … & Dean, J. (2021). Carbon emissions and large neural network training. arXiv preprint arXiv:2104.10350, p. 7. https://doi.org/10.48550/arXiv.2104.10350
Rivero, S., & Chinarro, D. (2024). CHANO: Inteligencia artificial como respuesta al colapso del Tribunal Constitucional y su impacto medioambiental como clave definitoria. Revista de Derecho Administrativo. Editorial Jurídica Sepín.
Rillig, M. C., Ågerstrand, M., Bi, M., Gould, K. A., & Sauerland, U. (2023). Risks and benefits of large language models for the environment. Environmental Science & Technology, 57(9), pp. 3464–3466. https://doi.org/10.1021/acs.est.3c01106
Sanh, V. (2019). DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter. arXiv preprint arXiv:1910.01108. https://doi.org/10.48550/arXiv.1910.01108
Schwartz, R., Dodge, J., Smith, N. A., & Etzioni, O. (2020). Green AI. Communications of the ACM, 63(12), pp. 54–63. https://doi.org/10.48550/arXiv.1907.10597
Sikora, A. (2021). European Green Deal: Legal and financial challenges of the climate change. ERA Forum, 21(4), pp. 545–558. http://dx.doi.org/10.1007/s12027-020-00637-3
Silva, S. R., & Vadillo, D. C. (2024). Inteligencia artificial y judicatura: sobre la dicotomía entre la asistencia y la sustitución. Aspectos técnicos y medioambientales. IDP. Revista d'Internet, Dret i Política, 41, pp. 1–12. https://doi.org/10.7238/idp.v0i41.426865
Sovacool, B. K., Monyei, C. G., & Upham, P. (2022). Making the internet globally sustainable: Technical and policy options for improved energy management, governance and community acceptance of Nordic datacenters. Renewable and Sustainable Energy Reviews, 154, 111793, 17. https://doi.org/10.1016/j.rser.2021.111793
Strubell, E., Ganesh, A., & McCallum, A. (2020, April). Energy and policy considerations for modern deep learning research. In Proceedings of the AAAI Conference on Artificial Intelligence, 34(9), pp. 13693–13696. https://doi.org/10.1609/aaai.v34i09.7123
Tarzanagh, D. A., Li, Y., Thrampoulidis, C., & Oymak, S. (2024). Transformers as support vector machines. arXiv preprint arXiv:2308.16898, p. 19. https://doi.org/10.48550/arXiv.2308.16898
Ulnicane, I. (2022). Artificial intelligence in the European Union: Policy, ethics and regulation. In The Routledge handbook of European integrations (p. 254). Taylor & Francis. http://dx.doi.org/10.4324/9780429262081-19
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention is all you need. In Proceedings of the 34th International Conference on Machine Learning (ICML 2017), Sydney. arXiv preprint arXiv:1706.03762. https://doi.org/10.48550/arXiv.1706.03762
Veale, M., & Zuiderveen Borgesius, F. (2021). Demystifying the Draft EU Artificial Intelligence Act. Computer Law Review International, 22(4), pp. 97–112. Retrieved from <https://ssrn.com/abstract=3896852> [Last consulted: January 18, 2025]
Wang, P., Zhang, L. Y., Tzachor, A., et al. (2024). E-waste challenges of generative artificial intelligence. Nature Computational Science, 4, pp. 818–823. https://doi.org/10.1038/s43588-024-00712-6
Wang, R., Zhang, X., Zhou, X., Wen, Y., & Tan, R. (2022, May). Toward physics-guided safe deep reinforcement learning for green data center cooling control. In 2022 ACM/IEEE 13th International Conference on Cyber-Physical Systems (ICCPS) (pp. 159–169). IEEE. https://dx.doi.org/10.1109/ICCPS54341.2022.00021
Zehner, O. (2012). Green illusions: The dirty secrets of clean energy and the future of environmentalism. University of Nebraska Press. https://doi.org/10.2307/j.ctt1d9nqbc
Zhang, P., Zeng, G., Wang, T., & Lu, W. (2024). TinyLLama: An open-source small language model. arXiv preprint arXiv:2401.02385. https://doi.org/10.48550/arXiv.2401.02385
Zheng, X., Pang, T., Du, C., Liu, Q., Jiang, J., & Lin, M. (2024). Cheating automatic LLM benchmarks: Null models achieve high win rates. arXiv preprint arXiv:2410.07137, 7. https://doi.org/10.48550/arXiv.2410.07137
Zhou, K., Zhu, Y., Chen, Z., Chen, W., Zhao, W. X., Chen, X., … & Han, J. (2023). Don't make your LLM an evaluation benchmark cheater. arXiv preprint arXiv:2311.01964. https://doi.org/10.48550/arXiv.2311.01964
Zhu, H., et al. (2023). Future data center energy-conservation and emission-reduction technologies in the context of smart and low-carbon city construction. Sustainable Cities and Society, 89(104322) http://dx.doi.org/10.1016/j.scs.2022.104322
Zhuk, A. (2023). Artificial intelligence impact on the environment: Hidden ecological costs and ethical-legal issues. Journal of Digital Technologies and Law, 1(4), pp. 932–954. https://doi.org/10.21202/jdtl.2023.40
