Atos is establishing a scientific committee in France tasked with guiding its thinking on open-source and sovereign artificial intelligence. This body is intended to advise the group on its technological choices and contribute to its innovation initiatives, notably those led within its Inno’Labs.
A team blending research, industry, and free/open-source software
The committee brings together four individuals from outside the group.
Alexandre Bouard, who leads the “connected vehicle” program of the Software République (a coalition of companies dedicated to mobility innovation) after fifteen years at BMW, will represent the automotive sector.
Stéphane Canu, professor at INSA Rouen Normandie and holder of a chair in statistical learning. He also serves as an IA policy officer at the Ministry of Higher Education and Research.
Philippe Freyssinet, scientific director of BRGM and member of the Academy of Technologies, is to shed light on the uses related to Earth sciences and the environment.
Finally, Jean-Baptiste Kempf, founder of the VideoLAN association behind the VLC media player, brings his experience from the open-source world.
Three Atos France executives complete the committee: Franck Chartier, Chief Executive Officer, Sylvain Barthélemy, Director of Technology and Operations, and Laurent Clergue, Director of Innovation.
A model more oriented toward open source
According to the IT services company, the committee should assist Atos in structuring an open AI foundation that brings together the group’s internal developments, with the aim of pooling work across teams rather than duplicating it. This approach should also feed an internal marketplace of AI agents, designed to facilitate the deployment of use cases at the group’s scale, especially in response to tenders related to digital sovereignty and the public cloud.
Beyond the industrial aspects, the committee will address several topics. The first concerns the use of AI in supporting patients and caregivers, particularly in the context of aging.
The committee will also examine the mathematical foundations of AI models, in order to better understand their biases and reliability limits; this facet will also include a scientific outreach dimension to young audiences, conducted in partnership with organizations recognized by the National Education system.
Finally, ethical questions will be part of the topics addressed, with a focus on model explainability, data protection, and the fight against algorithmic bias.
A committee aligned with the group’s AI offerings
This initiative follows two recent Atos announcements: Sovereign Agentic Studios, a mechanism described as aimed at moving sovereign-controlled agentic AI pilots into production, and MogwAI, the group’s generative AI platform.
