Digital Innovation & Energy Transition (EN)

Digital innovations:

challenges and opportunities

for the energy transition

EN | FR

Article written by Guillaume Bury for UNAPL IDF on May 20, 2025, without the help of an AI.

On Wednesday May 14, 2025, the Cercle de Giverny brought together experts, business leaders, and political representatives at the Maison de la Chimie in Paris to discuss the challenges of digital technologies in the era of the ecological and energy transition.

While digital innovations enable improved data production and analysis to accelerate the energy transition of our energy-intensive thermo-industrial society, these technologies are currently associated with increased electricity consumption and an ever-growing use of critical minerals, the demand for which could increase 3.5-fold by 2030 — a major asset for African countries (1). As we know all well, this has catastrophic consequences for natural ecosystems and biodiversity, as well as for the human rights of populations living in areas of mining exploitation (2).

Men and women from the public and private sectors sat down for roundtable discussions to address the digital challenges that concern French and European citizens:

Which digital responsibility exists in businesses?
How can AI support the energy transition?
How can we build digital infrastructure while limiting its carbon footprint within regions?
Can we claim digital sovereignty around cloud services?

Corporate digital responsibility

While limiting the carbon impact of data centers was already a topic of discussion, the massive expansion of generative artificial intelligence (AI) — such as ChatGPT, which now has over 1 billion users worldwide (3) — is not without its impacts on the planet and humans. It was recalled that the energy transition aims to expand the electricity market and reduce fossil fuels, particularly through the use of renewables and nuclear power. This is why the speakers agreed on the importance of optimizing electricity consumption and rethinking certain automated digital services in terms of their socio-economic added value — particularly given our dependence on their use for professional activities.

Nevertheless, there was also consensus on the desirability of digital technology in its ability to increase the competitiveness of French and European companies on the international stage, and thus establish our sovereignty. It is clear that we must strengthen the cybersecurity of public and private data against all attacks from both outside and within our borders. Faced with a near-vital need for regulation surrounding the use of AI and the operation of data centers, the European Commission proposed an initial legislation in April 2021 (4). Following this, the European Union adopted the AI Act in June 2024 — the world's first regulation on artificial intelligence.

How can this digital responsibility be put into action?

Sharing and adopting best practices among researchers and engineers.
Deploying progressive case law that goes beyond existing laws, which are regularly overtaken by the exponential (albeit limited) growth in AI system capabilities (5).
Supporting the creation of shared values as well as social and emotional skills — or soft skills — to capitalize on human capital, rather than increasing productivity when AI saves time in users' professional activities.
Preserving knowledge of the trade rather than promoting knowledge of AI prompts.

The job market is seeing technical skills become more autonomous, and must therefore be accompanied by an evolution of professional tasks. Let's leave data analysis to AI and work on human engineering and the soft skills necessary for the constant transformation of labour! In this context, young people in training or starting their professional careers must seize this opportunity to develop their abilities to adapt, communicate, be independent, engage in teamwork, reasoning, critical thinking, and creativity (6).

AI and the energy transition

In regards to the energy transition, the informed and rational use of AI promises significant optimisation potential in the following areas:

Reducing the gap between energy produced and the resources used to produce it.
Financial investment strategies for maintaining and improving hydrological and energy infrastructure.
Waste management techniques, including their reduction, recycling, and upcycling.
Developing energy efficiency models to increase our energy efficiency — the cleanest energy is the energy we don't produce!

This rational use aims first and foremost to prioritise traditional AI (e.g., machine learning and deep learning), which is already present within companies and operates on their own servers, as well as to limit unnecessary use of generative AI, which consumes significantly more energy — although its carbon footprint remains difficult to determine given the scale of its use (7). This would allow companies to better control their data and reduce their carbon emissions.

It is also about better predicting electricity consumption flows to limit losses and avoid unnecessary production. For these reasons, French public services and businesses must obtain information on generative AI currently operated by third-party companies of which they are clients, such as OpenAI or Microsoft, in order to understand and optimise their use.

Building sustainable digital infrastructure: is it possible?

Haute-Savoie MP Virginie Duby-Muller and OVHcloud representative Georges de Gaulmyn discussed environmental issues for areas hosting digital infrastructure — particularly data centers — and shed light on appropriate solutions. These include:

Promoting the multiple use of this infrastructure, with the example of creating heat networks to supply homes and local public buildings.
Prioritizing ethical stakeholders in public and private procurement. Using environmentally friendly cooling technologies (e.g., air cooling rather than water cooling).
Introducing greater transparency in digital services regarding the carbon trajectories of production, for a better understanding of environmental issues and informed decision-making regarding suppliers.

Cloud sovereignty and dependence

Akin to a sword of Damocles that threatens to strike us at any moment, as highlighted during Trump's major protectionist decrees following his second inauguration to the White-House on January 20, 2025, American data services still place European public and private actors in a situation of extreme dependence on the internet giants - Google, Apple, Facebook, Amazon, and Microsoft (GAFAM) (8). Faced with this distressing observation, the speakers emphasised the imperative of developing autonomy in hardware and software solutions, as well as in cloud services, all of which are essential to the activities of modern society.

We must lead by example: French public authorities at all levels must become aware of their digital dependence on GAFAM and reorient themselves towards French and European solutions, in order to set the course and regain our digital sovereignty!

References:

(1) Gourdon, J., Normand, E., Lapeyronie, H., et Bosse, P. (2024). Les minerais

critiques, des ambitions pour l’Afrique.

https://theconversation.com/les-minerais-critiques-des-ambitions-pour-lafr

ique-220735

Engérant, L., Barrieu, N. (2023). Les matières de l’immatériel : existe-t-il des

risques d’approvisionnement en matières premières pour les entreprises

du numérique ?. https://www.carbone4.com/analyse-risques-matieres-premieres-numerique

(2) Berthet, E., Lavalley, J., Anquetil-Deck, C., Ballesteros, F., Stadler, K., Soytas,

U., Hauschild, M., and Laurent, A. (2024). Assessing the social and

environmental impacts of critical mineral supply chains for the energy

transition in Europe. Global Environmental Change, 86, 102841.