French AI Research: February Synthesis

🧠 Traditional vs. AI-Generated Meteorological Risks for Emergency Predictions | SIRRI Naoufal, UMR6174 Institut Franche Comté Électronique Mécanique Thermique et Optique Sciences et Technologies (FEMTO-ST), Besançon, France. | Frontiers

Abstract: This study aims to analyze and examine in-depth the feature selection process using Large Language Models (LLMs) to optimize firefighter prediction performance. Although features from reliable sources are known to significantly aid predictions, their accuracy may be limited in critical situations requiring rigorous prioritization. Therefore, the focus was placed on meteorological risks for a comparative diagnosis between their extraction from M ét éo France and those generated by LLMs across various dimensions.

🧠 Bioethical challenges and artificial intelligence, focus Quebec/France | O. Gout, M. Lacroix. Ethics, Medicine and Public Health, Volume 33,
2025, 101058, ISSN 2352-5525.

Abstract: The authors discuss the potential benefits of AI for the healthcare system. To do this, they consider the importance of ensuring the confidentiality of medical data, maintaining a patient-doctor relationship imbued with humanity, as well as liability remedies specific to stemming the potential abuses of AI.

🧠 Where does AI come from? A global case study across Europe, Africa, and Latin America | Tubaro, P., Casilli, A. A., Cornet, M., Le Ludec, C., & Torres Cierpe, J. (2025). New Political Economy, 1–14.

Abstract: This article examines the organizational and geographical forces that shape the supply chains of artificial intelligence (AI) through outsourced and offshored data work. Bridging sociological theories of relational inequalities and embeddedness with critical approaches to Global Value Chains, we conduct a global case study of the digitally enabled organisation of data work in France, Madagascar, and Venezuela. The AI supply chains procure data work via a mix of arm’s length contracts through marketplace-like platforms, and of embedded firm-like structures that offer greater stability but less flexibility, with multiple intermediate arrangements. Each solution suits specific types and purposes of data work in AI preparation, verification, and impersonation. While all forms reproduce well-known patterns of exclusion that harm externalised workers especially in the Global South, disadvantage manifests unevenly in different supply chain structures, with repercussions on remunerations, job security and working conditions. Unveiling these processes of contemporary technology development provides insights into possible policy implications.

🧠 The Role of AI in Historical Simulation Design: A TPACK Perspective on a French Revolution Simulation Design Experience | Kindenberg, Björn. 2025. Education Sciences 15, no. 2: 192.

Abstract: This study explores the integration of generative artificial intelligence (GenAI), specifically ChatGPT, in designing a historical simulation of the French Revolution for eighth-grade students. Using the technological pedagogical content knowledge (TPACK) framework, the research examines how GenAI facilitated and obstructed the creation of an immersive educational experience, addressing the challenges and opportunities it presents. The study employs an explanatory case study methodology combined with autoethnographic elements, capturing the dynamic interplay between AI tools and educators in the design process. The simulation incorporated faction-based role-playing to engage students in historical decision-making, influenced by both pre-revolutionary and revolutionary events. GenAI played multiple collegial roles in the design process, including as a subject matter expert, game mechanics designer, and content communicator, enhancing efficiency and creativity. However, its limitations—such as unverified information, anachronisms, and biases—necessitated careful consideration, drawing on content matter expertise and knowledge of curriculum and class context. Findings indicate that the effective use of GenAI to assist simulation design requires a robust integration of content knowledge, technological proficiency, and pedagogical strategies within the TPACK framework. The study contributes to emerging research on AI’s role in pedagogical design process, with implications for history education and beyond.

🧠 Algorithm and Eve: how AI will impact women at work | Anon. 2024. OECD policy brief, 6 December 2024. Paris: OECD.

Abstract: In a recent study, female workers were 20 percentage points less likely to say they had used ChatGPT than male workers in the same occupation. While ChatGPT is just one [artificial intelligence] AI tool in a rapidly-evolving market, the finding raises questions about how women's and men's experiences of AI at work could differ. This is the question this policy brief aims to address, drawing from the OECD working paper 'Who will be the workers most affected by AI?' [available in VOCEDplus at TD/TNC 158.292]. The policy brief explores the gender composition of occupations highly exposed to AI and assesses women's access to AI-related employment opportunities and to productivity-enhancing AI tools. It concludes with a set of policy options that policymakers could pursue to ensure that women and men alike can benefit from AI at work.

🧠 Structuring and centralizing breast cancer real-world biomarker data from pathology reports through C-LAB ® artificial intelligence platform | Le Borgne F, Garnier C, Morisseau C, et al. DIGITAL HEALTH. 2025;11. doi:10.1177/20552076251323110

Abstract: To evaluate the effectiveness of C-LAB ® , an artificial intelligence (AI) platform, in extracting, structuring, and centralizing biomarker data from breast cancer pathology reports within the challenging, heterogeneous dataset of the Institut de Cancérologie de l’Ouest (ICO)

🧠 Improving care interactions (and training) in nursing homes with artificial intelligence | Lefelle, M., Samy Modeliar, M. GeroScience (2025).

Abstract: As the population continues to age, nursing homes will increasingly play a key role in caring for dependent individuals. To enhance the well-being of the elderly, it is crucial to focus on the language skills used during care interactions. However, issues such as the taboo surrounding dependency, scandals involving private nursing home management, the pressure for caregiver efficiency, and the variety of care contexts make monitoring these skills challenging. One way to address this is by collecting in situ data, supervised by language researchers and caregivers specialized in elderly care. This is the approach we have followed: the data collected was then analyzed using machine learning models to provide caregivers with crucial insights for improving care outcomes. Our research highlights the importance of specific factors in language-based interactions, especially in varied care situations. Notably, we emphasize the careful use of humor and the impact of caregiver experience on the success of care sessions. Consequently, we advocate for caregiver training that is grounded in real-life practice, focusing on context adaptation, active listening, and dialogue with residents.