At a time when generative AI is revolutionizing students’ research practices, CIVICA has launched an Information Literacy Barometer to better understand their habits, needs, and relationship with academic sources. In this interview, Sophie Forcadell, CIVICA’s European University Program Coordinator at the Sciences Po Library, and Cécile Touitou, Strategic Planning Coordinator at the Sciences Po Paris Library, discuss the key issues raised by this survey and the evolving role of university libraries.
Question 1: What is the CIVICA alliance? What are your responsibilities within CIVICA, and what led CIVICA to launch an Information Literacy Barometer among the alliance’s students today?
Sophie: CIVICA is a European alliance of social science universities, created as part of the European Union’s European Universities initiative. It brings together 10 universities to develop joint degree programs, support collaborative research, and promote mobility among students, faculty, and administrative staff. The alliance has been funded by the Erasmus+ program since 2019. The European Union’s goal is to build an integrated European higher education and research area focused on major societal challenges. Within the alliance, the partner libraries have organized themselves into a network and are part of a work package in which six working groups are implementing resource-sharing activities and services for the student and research community. My role at the Sciences Po library is to manage this aspect of the alliance project, from planning activities to reporting on them to the European agency in charge of the alliances.
Cécile: As the head of the Strategic Planning unit at the Sciences Po library and a specialist in user studies, I was responsible for leading a working group that oversaw the preliminary work leading up to the launch of the Information Literacy Barometer. When it came time to respond to the call for proposals (the Erasmus+ Alliance call), I suggested to Sophie that the CIVICA Alliance was ideally positioned to survey its student population about their information-seeking practices, their use of tools and resources, the hierarchy of sources they consulted, the reliability and credibility they attributed to them, and the role of university libraries—all just a few months after the widespread adoption of AI in their information landscape.
“Ultimately, I see university libraries evolving into the role of a trusted third party in matters of methodology—not in opposition to AI, but as guardians of robust information literacy that enables students to use AI in an informed and critical manner.”
Cécile Touitou
Foresight Project Manager at the Sciences Po Paris Library
Question 2: The report highlights a clear discrepancy between the starting points students prefer for their research (often search engines and generative AI) and the sources they consider most reliable (academic databases), as well as the fact that time constraints and access difficulties often lead them to take shortcuts. In your opinion, what is the first change that needs to be made to reconnect students’ daily practices with reliable academic resources?
Cécile: The situation is complex and certainly depends on the universities, the disciplines, and even the levels of study! In a recent article, we even learned that the sociocultural context is strongly correlated with the use of generative AI. In my view, the findings highlight the seriousness of the situation, particularly for undergraduate students, who are both the least experienced and the heaviest users of generative AI. It seems to me that, in this context, it would be essential for instructors to take action to encourage reading in class, reigniting a love of reading while emphasizing its academic value. To achieve this, it is important for students to understand what constitutes the quality and reliability of a text and to recognize the need to cross-check with primary sources. Indeed, one of the problems highlighted by the results is that, pressed for time and faced with multiple demands, students find it more convenient and efficient to rely on AI (and thus on content “pre-digested” by robots) than to engage in tedious reading. We must reverse this trend!
Workshops organized in partnership with libraries could be a valuable way to teach students how to read, research, and write for their academic work in the age of AI.
Sophie: These changes could be even more beneficial if they are designed at the level of the alliance’s library network, which already has its own AI best practices exchange group, bringing together more than 30 participants from the 10 libraries. The group plans to collaboratively develop a “train-the-trainer” program on AI, its recommended uses, and its limitations—a project largely informed by the survey’s findings.
Question 3: Given the ambivalence surrounding generative AI (widely used for support tasks but considered less reliable for source discovery and citation), as well as the high demand for information literacy training, how do you see the role of university libraries evolving in the coming years?
Cécile: The current context presents university libraries with a dual challenge: generative AI is revolutionizing academic research and writing practices, while the demand for information literacy training has never been greater, as the survey results show. These two trends reinforce each other, and this is precisely where libraries could play a central role.
It’s true that initiatives are still struggling to get organized, given the sheer number of projects. But a framework is beginning to emerge around a clearly identifiable chain of skills: information search, evaluating the reliability of sources, critical reading of content, and ethical use and citation. In each of these areas, librarians and teachers have complementary expertise, and it is the integration of these skills that could prove fruitful.
The risk would be that everyone works in isolation: librarians on information retrieval and citation, and teachers on subject-specific content, without these areas of knowledge feeding into one another. Conversely, a structured dialogue would allow the technical expertise of librarians to be integrated directly into teaching, particularly on issues where generative AI reveals its limitations: hallucinations, unverifiable sources, and bias in results.
Ultimately, I see university libraries evolving into the role of a trusted third party in methodology—not in opposition to AI, but as guarantors of robust information literacy that enables students to use AI in an informed and critical manner.
Depending on the topic—such as information searching, verifying the reliability of sources, analyzing content, and using and citing sources—faculty and librarians can position themselves within this chain of tasks that underlies all academic work. It would be a shame if libraries were to work in isolation on their areas of expertise (research training or source citation) without engaging in a fruitful dialogue with teachers to enrich their instruction with the technical expertise of librarians. But such collaborations are currently struggling to materialize. The framework of the alliance could make it possible to develop prototypes.
Sophie: The alliance framework allows us to collectively reflect on these types of questions, particularly within the group of partner library directors. It is clear that it is more necessary than ever to communicate very clearly that university libraries are there to support users in evaluating, contextualizing, and verifying information—and that this role is just as important as the long-standing one of organizing access to resources. Similarly, it is now very clear that the “information literacy” traditionally taught to students must be renamed “AI literacy,” at least for the foreseeable future. In my opinion, and given the legitimacy that the library group has gained within CIVICA’s governance structure, libraries can now develop institutional and alliance-wide recommendations on the use of AI, as well as support and guide academic integrity policies.
