National Green Algorithms Program: what it is, objectives, and benefits 

José Ragán

Director of ICT Operations – R&D&I Department · ACERTA CERTIFICATION

More than 5 years in software project certification and analysis of the impact of emerging technologies, such as AI, on the business world.

Artificial intelligence (AI) has become one of the main drivers of technological and economic transformation. However, its rapid deployment has also brought to light a growing challenge: the environmental impact associated with the intensive use of computational resources. It is in this context that green algorithms have emerged—an approach that seeks to minimize AI’s environmental footprint throughout its entire life cycle.

At the European level, this concern is part of a broader strategy aimed at reconciling technological leadership with sustainability and ecological transition goals. The European Union has begun to integrate criteria for energy efficiency, emissions reduction, and the responsible use of digital infrastructure into its digitalization policies, promoting artificial intelligence that is aligned with theEuropean Green Deal. Initiatives such as the European AI Strategy, the regulatory frameworks currently under development, and R&D&I funding programs encourage the design of more efficient, transparent, and sustainable algorithmic models, consolidating a common approach that seeks to reduce the environmental impact of AI without compromising its innovative capacity. 

More specifically in Spain, this vision is articulated through theNational Green Algorithms Program, a pioneering initiative that promotes the development ofsustainable artificial intelligence, integrating environmental criteria from the design phase of algorithmic systems. 

What are green algorithms and why do they matter? 

Green algorithmsare those designed toreduce their environmental impact, taking into account not only their functional performance, but also the energy consumption and resources required for their training, deployment, and maintenance. 

The rise of increasingly complex AI models has significantly increased the energy consumption associated with training and inference. This phenomenon is exacerbated when models are not optimized, are trained redundantly, or are run on energy-inefficient infrastructure. In light of this reality, green algorithms aim to: 

• Optimize the use of computational resources. 
• Reduce energy consumption during training and inference. 
• Minimize emissions associated with the use of hardware and data centers. 

Their importance lies in the fact that they enable technological progress to be reconciled with sustainability, aligning the development of AI with ESG (environmental, social, and governance) principles. 

The need for sustainable artificial intelligence

The widespread adoption of AI in virtually all productive sectors has brought significant benefits in terms of efficiency, productivity, and innovation. However, it has also highlighted the need to incorporate environmental criteria into technological development.

From a sustainability perspective, it is not enough to analyze the impact of AI during its operational phase. It is essential to consider its entire life cycle, from algorithm design to deployment and updates. In this regard, a lack of measurement and control of energy consumption can lead to significant environmental impacts.

Green algorithms address precisely this need, promoting AI that considers environmental impact as just another design variable, on par with accuracy or functional efficiency.

The magnitude of this challenge becomes clear when analyzing the energy consumption associated with training large-scale artificial intelligence models, particularly in the field of deep language models. 

Widely cited estimates place the energy consumption of training models such as GPT-3 in the order of thousands of MWh, a figure comparable to the annual electricity consumption of dozens or even hundreds of households. Although these figures are approximations, they show that decisions made during the design, training, and optimization phases of algorithms have a direct and significant impact on energy consumption. In this context, the incorporation of strategies aimed at computational efficiency and the reduction of algorithmic complexity is no longer a purely technical issue, but has become a key element of sustainability and technological responsibility. 

What is the National Green Algorithms Program?

The National Green Algorithms Program is a Spanish initiative aimed at promoting the development of sustainable artificial intelligence by incorporating environmental criteria from the earliest stages of the design, development, and deployment of algorithmic systems.

The program aims to promote AI that is "Green by Design," in which energy efficiency, emissions reduction, and the responsible use of digital infrastructure are integrated as key technical variables, alongside model accuracy and performance.

Within this framework, the National Green Algorithms Program acts as an instrument for aligning technological innovation and sustainability, promoting good practices, measuring the environmental impact of AI, and developing standards and tools that facilitate its implementation in companies and organizations. 

Objectives of the National Green Algorithms Program

TheNational Green Algorithms Programwas created with the goal of promoting AI that is green by design, encouraging the development of algorithms that reduce their environmental footprint from the earliest stages of the process.

Among its main objectives are:

• Integrate environmental criteria into the design of AI algorithms and models.
• Promote best practices in energy efficiency and computational optimization.
• Promote the measurement of the environmental impact associated with the development of AI.
• Promote the adoption of standards and frameworks for digital sustainability.

This program is part of a broader strategy

Measuring to improve: the importance of the digital carbon footprint

One of the key principles of green algorithms is that you cannot reduce what you do not measure. Therefore, measuring the energy consumption and emissions associated with AI systems is a fundamental step. 

There are already initiatives that allow the carbon footprint of models and algorithmic systems to be estimated, taking into account factors such as: 

• Hardware energy consumption. 
• Data center efficiency. 
• Indirect emissions associated with technological infrastructure

This information is essential for identifying opportunities for improvement and making informed decisions in the development and deployment of AI solutions. 

Strategies for developing green algorithms 

The development of green algorithms involves the application of various technical strategies, including: 

• Neural network optimization and complexity reduction. 
• Efficient use of hardware and selection of energy-optimized infrastructure. 
• Reuse of previously trained models and weights. 
• Reduction of unnecessary training and redundant processes. Accurate knowledge of your climate impact.  

These practices not only contribute to reducing environmental impact, but can also generate economic savings and improvements in operational efficiency. 

In other words, the development of green algorithms responds to the need to rethink traditional models for designing and training artificial intelligence systems, incorporating efficiency and sustainability criteria as top-level technical variables. Compared to approaches based exclusively on increasing computing power and data volume, this approach promotes optimized architectures, more selective training processes, and the rational use of available computational resources. In this way, sustainability ceases to be an external factor and is directly integrated into the engineering decisions of the algorithmic system itself. 

From an operational and strategic perspective, adopting these practices allows us to move toward AI solutions that are more robust, scalable, and economically viable in the long term. Reducing energy consumption, decreasing computing times, and optimizing the use of technological infrastructures contribute not only to mitigating environmental impact, but also to improving the maintainability, reproducibility, and overall efficiency of the systems developed. In this sense, green algorithms are establishing themselves as a key approach for the development of responsible artificial intelligence that is aligned with current technological challenges. 

How to reduce your carbon footprint once it has been calculated 

The adoption of green algorithms brings multiple benefits to organizations, both from an environmental and strategic point of view. Among them are: 

• Reducing the digital carbon footprint . 

• Optimization of energy consumption and associated costs. 

• Improvement of corporate reputation and alignment with criteria  

• Anticipating future regulatory requirements in technological sustainability. 

Furthermore, incorporating sustainability criteria into AI development strengthens the trust of customers, investors, and other stakeholders. 

The role of certification and reference frameworks 

The National Green Algorithms Program also promotes the development of certification tools that enable objective assessment of the degree of sustainability of AI systems. 

In this area,ACERTA participatesin the development of certification schemes aimed at ensuring that algorithms meet environmental criteria from the outset, thereby contributing to greater transparency and confidence in the use of artificial intelligence. Certification becomes a key element for: 

• Verify compliance with sustainability criteria. 
• Compare technological solutions objectively. 
• Promote best practices in the digital ecosystem. 

The introduction of specific certification frameworks for green algorithms makes it possible to translate sustainability principles into measurable, verifiable, and comparable indicators, facilitating their integration into technology development and evaluation processes. These certification frameworks help to objectively assess aspects such as energy efficiency, computational optimization, and responsible algorithm lifecycle management, providing organizations with reliable tools to evaluate the environmental impact of their artificial intelligence systems in a systematic and reproducible manner.

From a strategic perspective, certification in algorithmic sustainability thus serves as a driving force for continuous improvement and the adoption of advanced standards in the digital ecosystem. By providing clear and measurable criteria, it fosters the development of more efficient and responsible solutions, while strengthening the trust of users, customers, and government agencies in the use of artificial intelligence. In this regard, the certification not only validates compliance with environmental requirements but also promotes a model of technological innovation aligned with the principles of transparency, efficiency, and sustainability. 

A vision for the future of sustainable AI 

Progress toward sustainable artificial intelligence requires a collaborative approach involving government agencies, companies, research centers, and technical organizations. The National Green Algorithms Program represents an important step in this direction, positioning sustainability as a central pillar of technological development.

Integrating environmental criteria into AI is not only a responsibility but also an opportunity to build a more efficient and transparent model of innovation that is aligned with today’s global challenges.

This collaborative approach enables progress toward common standards, shared methodologies, and best practices aimed at measuring, optimizing, and reducing the environmental impact of artificial intelligence systems. Coordination among the various stakeholders facilitates knowledge transfer, the validation of sustainable technological solutions, and the gradual adoption of development models that prioritize computational efficiency from the earliest stages of algorithmic design. In this way, sustainability is established as a cross-cutting criterion in the evolution of AI, moving beyond isolated or one-off initiatives.

In this context, green algorithms are emerging as a key element in ensuring the long-term viability of artificial intelligence, especially in scenarios involving scaling and widespread adoption. The systematic incorporation of environmental criteria helps reduce operational and energy risks, improve technology governance, and strengthen confidence in the use of AI as a tool for economic and social transformation. Thus, sustainability not only accompanies technological development but also becomes a structural factor in competitiveness and responsible innovation. 

Conclusion

Green algorithms areestablishingthemselvesas a key element in advancingtoward sustainable artificial intelligence, capable of generating economic value without compromising the environment. The National Green Algorithms Program paves the way for incorporating sustainability into the design of algorithmic systems, promoting responsible innovation aligned with environmental objectives. 

The future of AI necessarily involves integrating efficiency, transparency, and sustainability into every stage of its development.