Published on May 22, 2024

La Poste strengthens the ethical design of its artificial intelligence projects with a new data & AI charter

As part of its innovation strategy, La Poste wishes to seize the opportunities offered by artificial intelligence (AI) by developing “ethics by design” projects, consistent with its values of trust and integrity.

To achieve this goal, La Poste Group is adopting a new data & AI charter to govern the design and development of artificial intelligence projects, including those incorporating generative AI.

The ethical principles set out in La Poste Group's Data & AI Charter define an operational framework for action geared towards the reasoned use of data and the ethical design of algorithms. Compliance with the charter's commitments is ensured by the Committee for Trusted AI, made up of internal and external experts. This approach marks a new milestone in the Group's maturity and the industrialization of trusted AI.

Nearly 100 data and AI projects have been implemented or are under development across the Group. Among them, La Poste is presenting the La Poste GPT demonstrator, a generative AI solution developed in-house.

A data & AI charter consolidating the commitments made in 2016

To support the growing use of artificial intelligence (AI) and generative AI by its 500+ experts, La Poste Group has adopted a new data & AI charter, applicable to all projects developed in its branches and subsidiaries. Published on the occasion of VivaTech, this charter sets out clear guidelines for the reasoned use of data and the ethical use of artificial intelligence, in keeping with La Poste's position as a trusted third party. This charter and the accompanying application framework guide La Poste's choices in the design and development of trusted, ethical and responsible solutions. The scheme aims to implement AI projects that are respectful of fundamental rights, transparent in their operation and guarantee a responsible approach, in order to raise awareness and prepare for the entry into force of the European regulation on artificial intelligence.

This new charter consolidates the commitments made back in 2016 with an initial “data charter”, anticipating the General Data Protection Regulation (GDPR), supplemented in 2021 by the “ethical principles and artificial intelligence”, which anticipated the application of the future European regulation on AI.

A committee of experts to monitor compliance with commitments

To ensure compliance with the principles of this new charter, La Poste has set up the Comité pour une IA de confiance (Trusted AI Committee), made up of Group experts and qualified external figures. Its main missions are to ensure compliance with the Group's deontological and ethical principles, to support project teams in determining data processing purposes, and to guarantee the ethical process in project development.

The committee has 23 members, including 5 external experts :

  • Françoise Soulié-Fogelman, scientific advisor to the Hub France IA and artificial intelligence expert to the European Commission

  • Vincent Roca, researcher, in charge of the PRIVACTIS project at INRIA

  • Éric Dadian, President of the French Customer Relations Association (AFRC)

  • Célia Zolynski, professor of private law and coordinator of the AI Observatory at Université Paris 1 - Panthéon-Sorbonne

  • Eric Salobir, Chairman of the Executive Committee of the Human Technology Foundation (HTF)

Training employees in the challenges of trusted AI

The Group's new data & AI charter reinforces La Poste's ethical commitments in the use of artificial intelligence. It sheds light on our choices in the design, processes and service offerings using AI. With this charter and the committee for trusted AI that ensures compliance with its principles, La Poste aims to position itself as an exemplary player in the ethical and reasoned use of the possibilities offered by this disruptive technology.

Pierre-Etienne Bardin

Chief data officer of La Poste Groupe

With the aim of acculturating and supporting the Group's 230,000 employees in the challenges of trusted AI, awareness-raising training courses are made available to them. Advanced training courses are included in the training program for data-related professions.

In addition, and as part of its trusted AI approach, La Poste is participating in the development of the Afnor SPECS frugal AI standard in partnership with the French Ministry of Ecological Transition and Territorial Cohesion, in order to build environmentally-friendly AI systems.

With more than 100 projects under development in its branches and subsidiaries, La Poste is experiencing strong momentum around the uses of artificial intelligence. At the next edition of VivaTech (May 22 to 25), the group will present La Poste GPT, a generative AI developed within technological environments mastered by the group and trained on its own databases to improve performance. Among the first uses identified: supporting sales teams in responding to their customers' needs and assisting customer relations teams to gain time and relevance in resolving requests. More than 200 group employees are currently testing the solution, for which a gradual rollout is planned for the end of the year.

Read more