- The UNE 0086 specification establishes a common framework for assessing the energy consumption, carbon footprint, and resource use of Artificial Intelligence models.
- In this way, both organizations are promoting a pioneering standard to achieve more sustainable and responsible AI, developed with input from over 70 experts representing 35 public and private entities.
Madrid, 6 November 2025 – The Spanish Association for Standardization, UNE, has published the new technical specification UNE 0086, a document developed in collaboration with organizations such as Adigital, which establishes a common framework for measuring the energy consumption, carbon footprint, water usage, and performance of AI systems.
Currently, data centres consume 60 TWh of electricity annually, equivalent to the total annual consumption of the European rail network, and this figure could double by 2030 according to the International Energy Agency (IEA). Aware of this situation, the Secretariat of State for Digitization and Artificial Intelligence has promoted the standard within the framework of the National Green Algorithms Programme (PNAV).
The standard was developed with the collaboration of more than 70 experts from 35 entities, including Adigital, aiming to provide proposals focused on transparency and communication of these indicators so they are more understandable and easily applicable by companies, professionals, and the public.
“In line with our goal of advancing human-centred, transparent, and sustainable digitization, at Adigital we wanted to provide a practical perspective on how to communicate and apply sustainability criteria in the use of artificial intelligence. As representatives of Spain’s largest technology business community, we believe this specification will be a key tool to drive future certifications that reinforce trust in responsible, ethical AI aligned with European values”, says Justo Hidalgo, Director of Artificial Intelligence at Adigital
Pioneering standard
The UNE 0086 specification provides detailed guidance for quantifying the environmental impact of AI models and algorithms across all stages of their lifecycle, with particular attention to generative models and large language models (LLMs) due to their high computational demands. It includes key indicators to monitor the direct use of energy and water, as well as the overall efficiency of models during training and inference.
“With this specification, UNE strengthens its commitment to creating standards that contribute to more sustainable and responsible digitization. UNE 0086 is an example of how standardization can anticipate technological challenges and offer concrete tools for organizations to measure and reduce their environmental impact when using Artificial Intelligence”, notes Javier García Díaz, Director General of UNE and Vice President of ISO.
This initiative is part of the 2024 National Artificial Intelligence Strategy and reflects the Spanish Government’s commitment to ethical, transparent, and sustainable AI. Moreover, it represents a decisive step towards future European standardization, aligning with the Code of Good Practice for general-purpose AI models published by the European Commission in August 2025.
With this specification, Spain positions itself at the forefront of defining international standards in technological sustainability, reinforcing its role as a benchmark in the development of artificial intelligence in service of collective progress.
