Sustainable and digital transformation: an inevitable convergence in the age of AI?
Published July 2, 2025
Key takeaways
- Shared use cases between IT and sustainability teams are expanding and becoming increasingly complex.
- AI holds great promise for advancing sustainability strategies, but it also raises growing concerns about its impact.
- There is broad agreement on the need for AI to be both efficient and responsible, respecting planetary boundaries and fundamental rights.
- Embedding sustainability across all business functions and ensuring robust data governance are paving the way for new forms of collaboration.
Digital and sustainability: use cases are multiplying and growing more complex
Collaboration between sustainability teams and digital & IT departments is not entirely new in large organizations, particularly when it comes to measuring and managing non-financial performance. However, needs are diversifying and extending well beyond the sustainability function and its non-financial reporting mandate, impacting procurement, risk management, supply chain, marketing, and more.
IT departments now need to ensure that sustainability data is seamlessly integrated across all elements of their business information systems. This trend is mirrored in the software market, which remains highly fragmented despite some initial consolidation. No single SaaS solution can cover every need, underscoring the importance of setting clear roadmaps for the next three to five years and targeting the right investments.
A few examples
Crédit Agricole has been developing its sustainability tools for years, well before the CSRD, to meet the banking sector’s regulatory requirements for non-financial reporting. Today, its sustainability strategy aims to reach every part of the organization, integrating with digital capabilities to enhance ESG insights on clients and to offer more tailored products.
La Poste is seeing new needs emerge around climate change adaptation, resource and vulnerability management, and measuring the carbon footprint of its customers.
EDF aims to help build a carbon-neutral future through innovative solutions and services where digital technologies play a central role, from IoT and data to AI. This translates into tools that help consumers manage their energy use and into remote maintenance that keeps industrial assets operating longer and more efficiently.
Barely 10% of companies worldwide currently measure their emissions comprehensively. Given the complexity of the data to be processed, digital technology has become indispensable for precisely calculating the carbon footprint and managing decarbonization trajectories
Sustainability must be embedded by design into all business lines and functions across the Group, and integrated with digital tools — particularly to deepen our ESG understanding of clients.
The CSRD is accelerating the convergence of digital and sustainability.
AI holds great promise but raises growing ESG concerns
The launch of ChatGPT at the end of 2022 marked a turning point in the rise of generative AI, which has quickly been adopted across nearly every business function. Sustainability teams, already focused on building data capabilities to meet regulatory demands and deliver on voluntary commitments, are now weighing the opportunities and risks that come with these technologies.
At the heart of the discussion: how can companies balance the growth of generative AI with sustainability goals? While AI raises major ESG questions — particularly regarding the energy and resource demands of its infrastructure and the related ethical implications — certain use cases show real potential to deliver non-financial value for businesses.
A few examples
At La Poste Group, AI and data are strong business drivers. Docaposte, its digital trust subsidiary, supports healthcare professionals (via Dalvia Santé) in managing and caring for patients. Meanwhile, Geopost, the express parcel delivery arm, is developing autonomous electric delivery robot solutions.
CO2 AI, a carbon accounting software provider, assists Reckitt (an Anglo-Dutch multinational specializing in cleaning and hygiene products) in its decarbonization efforts across its entire value chain (Scope 3). Using its technology, CO2 AI enables the association of the most relevant emission factor with each product line in just a few minutes, automating a task that was formerly manual, time-consuming, and prone to errors. The tool accelerates analysis and enhances the accuracy of the calculations necessary for the group’s low-carbon transition.
Thanks to predictive modeling, AI helps optimize the replacement of major components in nuclear power plants, extending their lifespan and reducing their carbon footprint
Augmented logistics helps reduce empty miles, aligning ecological impact with economic efficiency.
Augmented logistics helps reduce empty miles, aligning ecological impact with economic efficiency.
Consensus around a responsible and resource-efficient AI
Despite the continued lack of transparency from hardware manufacturers and software providers about the true environmental impact of generative AI, organizations are working to factor environmental as well as social and societal considerations into how they develop and deploy artificial intelligence. This commitment to managing impacts is reflected in multiple initiatives across the entire value chain, with a clear ambition to make AI more efficient and responsible — and with special attention paid to ethical and inclusive practices.
A few examples
La Poste has established an AI charter built on three principles: transparency, protection of fundamental rights, and ethics. It is putting this commitment into practice in its IT systems by choosing less energy-intensive data centers. La Poste has also launched its own School of Data and AI, aiming to train up to 250 candidates per year over the next three years in four key roles: data product owner, data analyst, data engineer, and data scientist.
At EDF, teams are working to embed eco-design principles into the development of AI agents. The preferred models are smaller language models (known as Small LLMs), sometimes used with RAG*, as they are lighter, less resource-intensive (in terms of memory and energy), and faster to run, in some cases even operating locally.
An effective and meaningful AI strategy must be a resource-efficient one — not just for environmental reasons, but for economic ones too. Today’s societal challenges also include sovereignty, which must now be treated as an integral part of ESG. From this perspective, AI represents a real challenge.
An effective and meaningful AI strategy must be a resource-efficient one — not just for environmental reasons, but for economic ones too. Today’s societal challenges also include sovereignty, which must now be treated as an integral part of ESG. From this perspective, AI represents a real challenge.
New forms of collaboration
The challenges linked to the responsible use of AI and robust ESG data governance are placing IT departments at the heart of the sustainable transformation. Beyond leading responsible digital initiatives, CIOs are now expected to help shape IT master plans that reflect a holistic view of ESG priorities across all business functions. This shift is paving the way for new forms of collaboration and organizational models.
A few examples
At Crédit Agricole, a new Transformation unit has been created, bringing together sustainability, digital, and HR. This cross-functional approach is enabling closer collaboration between ESG and digital teams.
Within La Poste Group, the CSRD has served as a catalyst for significant changes in structure and sponsorship, acting as a real accelerator of mutual understanding and collaboration between the sustainability and IT departments. A key objective is to develop a shared, systemic vision of ESG integration.
Digital and ESG are both powerful drivers of business transformation. Their alignment and joint implementation are essential levers for delivering real value to customers and teams
It’s essential not to work in isolation. Joining communities like INR helps us learn faster by sharing feedback and best practices. At INR, we co-create practical tools, frameworks, training programs, and certifications — all of which help organizations boost their impact
Wavestone sincerely thanks the CSR and IT leaders who shared their experiences and beliefs
- Claire Baritaud, Director of Policy, Expertise and Planning, Societal Engagement Department, La Post
- Richard Bury, Head of the Responsible Digital Program, EDF – President, Institut du Numérique Responsable (INR)
- Charlotte Degot, Founder & CEO, CO2.AI
- Christophe Jacolin, Head of ESG Strategy, Crédit Agricole
*RAG (Retrieval-augmented generation) allows enriching LLM data with those of the organization (the amount of data needed remains low, and this practice avoids the costs and delays of fine-tuning or pre-training). More information here: https://www.databricks.com/fr/glossary/retrieval-augmented-generation-rag
