Érudit's RSS Feed

Les milieux documentaires face à l’IA

Réjean Savard

L’intelligence artificielle générative en bibliothèque universitaire : évolution des pratiques de recherche documentaire et du rôle des bibliothécaires

Mylène Raymond, B. Sc., M.B.A and Jason Luckerhoff, Ph. D., Adm.A., CRHA
This research documents the transformations brought about by generative artificial intelligence in Quebec university libraries. It is based on 118 interviews with librarians, professors, and students, analyzed using grounded theory methodology. Eight themes emerge from the data, covering practices, professional roles, user needs, staff concerns, and governance issues. The analyses show that generative artificial intelligence is not perceived as a disruption, but as a transformation in continuity with the digital shift already underway. Documentary work is shifting from identification to validation: the linguistic plausibility of the responses generated can mask errors, biases, or non-existent references. The efficiency gains perceived upstream are often absorbed by downstream verification. The role of librarians is being redefined around artificial intelligence literacy, critical mediation, and organizational structuring. On the user side, a tension is emerging between the need for flexible access and the link to human support. Institutional oversight remains uneven: uses are developing faster than policies.

Pratiques informationnelles avec l’IA en bibliothèques : enjeux et appropriations

Emmanuelle Chevry Pébayle and Arnaud Zeller
Artificial intelligence (AI) is now a major technological lever for libraries, with growing applications in cataloguing, mediation, conservation and strategic management. Recent advances, particularly in generative AI, have made it possible to automate complex tasks such as generating records or classifying images, profoundly transforming professional practices. However, the literature highlights concerns and a persistent conceptual vagueness around the use of AI, giving rise to uncertainty and questioning among professionals and decision-makers. Against this backdrop of change, libraries need to rethink their missions and anticipate the implications of these technologies. The aim of this exploratory study is to examine the perceptions of professionals regarding AI, to identify the challenges they face and to document the first uses of these technologies in user services. The analysis falls within the theoretical framework of social and professional representations.

L’intelligence artificielle au service de la découvrabilité des archives historiques : quelles parties prenantes et quelles collaborations ?

Siham Alaoui, M.S.I., Ph. D.
Cultural institutions are increasingly adopting technologies such as artificial intelligence (AI) tools to improve their practices to manage collective memory. In the archival sector, the use of these technologies supports the processing, dissemination, and preservation of historical archives. While the integration of AI into archival processes offers considerable gains in terms of resources optimization, it also carries some risks. Thus, the automation of archival processes requires clear guidelines governing the use of disruptive technologies and close collaboration between multiple stakeholders with different objectives. This contribution aims to address the prospects for integrating AI into the processing and dissemination of archives, while emphasizing the importance of the collaboration between various stakeholders to enhance the quality of these historical records.

Réinventer les services : BAC mise sur l’IA et la technologie pour mieux servir le public

Sarah Bellefleur Bondu, M.S.I. and Catherine Butler, M.A.
Library and Archives Canada is undertaking a comprehensive digital transformation centered on user experience, modernizing a wide range of services offered both onsite and online. The institution is investing in both traditional tools (search interfaces, the personalized LAC Account space, online forms, new access to information platforms) and transformative technologies such as artificial intelligence, notably through pilot projects with Transkribus and the automated transcription of the 1931 census. This evolution is supported by a dedicated user experience team, the adoption of agile working methods, service blueprinting, and a feedback framework that systematically incorporates the voices of users and staff. In a governmental context characterized by limited resources and high expectations, Library and Archives Canada is implementing a well-governed approach to artificial intelligence, aligned with the Policy on Service and Digital and with its Vision 2030, while preparing the next steps: a new client relationship management system, conversational channels, and targeted automation, in anticipation of the opening of Ādisōke, the new primary service point of Library and Archives Canada in Ottawa.

L’IA au service de la recherche généalogique au sein des bases de données de presse entre enjeux algorithmiques et épistémologiques : le cas de RetroNews et de France, noms et histoires dans les journaux

Victoria Laurent and David Francisco
Are public online newspaper collections on the verge of becoming commercial commodities? This is one of the questions addressed in this article, which examines the changes that have taken place in the field of genealogical research following a wave of digitization of French newspapers since the turn of the millennium. This process is linked to the emergence of private platforms forming partnerships with public heritage collections. Significant changes are thus observed in how these databases are utilized and in the ways these documentary collections are searched by those we have chosen to call “ancestronauts”. Furthermore, the proliferation of algorithmic tools in their processing has led to a profound shift in the balance, creating various biases that we will attempt to identify in this paper, focusing on RetroNews and France, noms et histoires dans les journaux, based on an analysis of platform interfaces and design coupled with a qualitative study.