AI and Sustainability: Turning Technological Innovation into a Driver of Performance

The integration of artificial intelligence is entering a pivotal phase for businesses. At Tennaxia Connect, Rémy Balangué, Chief Product Officer at Tennaxia, and Grégoire Carpentier, Chief Technology Officer at Tennaxia, shared their perspectives on the challenges of applying AI to sustainability. Moderated by journalist Sébastien Borgnat, this session provided an opportunity to compare technological promises with operational and environmental imperatives. Covering topics ranging from optimizing data processing and managing the digital carbon footprint to governance requirements, the conference demonstrated that AI must not be merely a publicity stunt. When properly integrated, it becomes a clear, useful framework aligned with organizations’ goals for sustainable performance and strategic management.

Why AI and Sustainability Are Converging Now

The intersection of artificial intelligence and sustainability initiatives comes at a time when the volume of data to be managed—whether regulatory texts, charters, or non-financial reports—is becoming critical for companies. To gauge the reality on the ground, Tennaxia conducted a large-scale survey of its clients. The results highlight three particularly telling figures:

• 60% of respondents already use AI systems as part of their work, but through personal accounts on consumer-facing platforms. AI has thus found its way into the workplace even before being formally integrated by IT departments (CIOs).
• 70% of them cite time savings as the primary benefit they expect from AI in their daily work.
• 20% say they do not use it at all, mainly due to a lack of confidence or because they do not yet consider the cost-benefit ratio to be satisfactory.

This shift confirms that the primary focus is no longer on discovering the technology, but on the ability to manage it. Companies are seeking solutions that can deliver real productivity gains while ensuring a trustworthy environment.

A pragmatic approach that prioritizes real-world usefulness over gimmicks

The integration of AI into CSR and EHS initiatives should EHS aim to disrupt everything simply for the sake of innovation. As Tennaxia points out , the goal is to focus on real business needs to avoid “gimmicky” tools. Large language models (LLMs) are, by nature, text processors designed to predict output text based on input text. They are therefore inherently well-suited for managing sustainability data, which relies heavily on text and tabular formats.

The goal is not to reinvent existing systems, but to make them more efficient where the administrative burden is heaviest: searching for information, rephrasing content for different audiences, or consolidating scattered sources. AI is most effective when it informs decisions without complicating business processes.

The three key features launched in the ESG area

To put this approach into practice, Tennaxia has rolled out three specific features on its ESG platform this year:

• Bulk questionnaire translation: This feature allows you to automatically translate a reference questionnaire into more than 20 languages to facilitate data collection from global subsidiaries without leaving the application.
• Creating dashboards using natural language: The user formulates a request directly to automatically generate the desired dashboard, complete with the appropriate metrics, data, and graphical representations.
• Support for drafting the sustainability report: a module capable of consolidating and reusing information from previous reports or policy documents to pre-fill data for the current fiscal year.

Internal Pilot Programs and the Transformation of EHS Functions

In the EHS area, the approach relies primarily on internal testing prior to large-scale deployment. Tennaxia is currently testing AI tools for audits with its own consultants:speech-to-text for transcribing meeting minutes, note-taking via photos, and the automatic generation of summary tables or presentation materials.

This approach helps validate the tool’s functional relevance and standardize practices early on. By converting writing time into proofreading time, AI acts as a catalyst. Above all, it changes the nature of tasks by paving the way for new decisions.

Real-world example: When opening a factory in a new country, the analysis of unfamiliar and voluminous regulations—a task often so tedious that it can delay a project—can now be facilitated and accelerated by automated reading and summarization capabilities, thereby enabling decision-making.

The Environmental Footprint of AI: The Path to Sustainability

One of the conference’s key messages concerns the environmental responsibility associated with the use of technology itself. AI is a major consumer of energy, a critical issue for sustainability stakeholders. Globally, data center energy consumption totals more than 400 TWh. AI currently accounts for 15% of this consumption, but according to the International Energy Agency (IEA), this share could reach at least 30% by 2030.

In light of this observation, management must incorporate criteria for digital restraint, drawing in particular on the work of organizations such as Green IT and the think tank The Shift Project. Tennaxia applies this approach through two key strategies:

• An internal usage policy supported by ongoing scientific monitoring to track key publications.
• Common-sense heuristics during development, which involve choosing the most appropriate and efficient model for a given task, rather than systematically relying on the most resource-intensive and energy-consuming infrastructure.

Trust, Sovereignty, and the Centrality of the Human Being

The reliability of AI remains a major concern, as no model is 100% reliable. The solution to this limitation lies in strict governance based on the principle of "human-in-the-loop" (the human in the loop). On the platform, AI operates exclusively in “opt-in” mode: it is never enabled by default and requires explicit action by the company. Furthermore, the tool is limited to suggesting text or pre-filled fields; it cannot overwrite or replace data without explicit, fully traceable, and time-stamped human validation.

In terms of sovereignty and security, data is stored and processed within the European Union—specifically in France—to ensure the highest levels of confidentiality (which require that data be hosted and processed in French data centers). Customer data is never used to train suppliers’ models.

Finally, this approach aligns with ISO 27001 and SOC 2 requirements, ensuring the system’s auditability and resilience: the software is designed to operate without AI, thereby preventing any loss of functionality for the company in the event of a technical outage or a sudden surge in technology costs. This architecture is also aligned with the requirements of the European AI Act, whose provisions will be phased in gradually through 2027.

From Technology to Sustainable Management

Artificial intelligence does not require companies to develop technology for its own sake, but rather to use it more effectively to make better decisions. In an environment where sustainability and compliance requirements are becoming increasingly stringent, it provides a useful framework for moving from reactive data collection to strategic management. By automating repetitive administrative tasks, it restores value to human expertise on the ground.

For organizations, the challenge now is to turn enthusiasm for technology into an opportunity for transformation: to make AI a well-managed, streamlined, and autonomous foundation for driving CSR performance that is robust, consistent, and fully manageable.