Article by Reece Steinberg, head, Library Learning Services
Originally published in Academia Forum, March 9, 2026
The recent conversation around generative AI in higher education has often been defined by anxiety. For many instructors and administrators, the primary concern remains academic integrity and the potential for students to use these tools to bypass original thought. However, as I recently discussed in an interview with CBC, focusing on policing limits our ability to support students in AI experimentation and skills development. At Toronto Metropolitan University (TMU) Libraries, we have decided to shift the focus from use to fluency.
The distinction between literacy and fluency is important. While literacy suggests a basic understanding of what a tool is, fluency implies the ability to use that tool effectively, ethically, and transparently. To address this, TMU Libraries launched a series of AI fluency workshops with badges (credentials students can share on social media). These sessions are designed to help students understand how and when to approach large language models and the ethical implications of their use in research.
The AI “Black Box” Problem: A Library Issue
One of the most significant challenges in the current landscape is the black box nature of AI: Users can’t easily trace where claims come from or how an output was produced. Students often use these tools without knowing where the information originates or what sources underpin the information they receive. This makes it harder to verify accuracy, assess bias, or cite responsibly.
Librarians are the natural leaders for this initiative because our core mission has always been about information literacy. We have spent decades teaching students how to evaluate sources, cite information, and navigate complex databases. Generative AI is the latest evolution of the information landscape. The same critical thinking skills required to spot a predatory journal or a biased news article are required to evaluate an AI-generated summary.
During the development of our badging workshops, we found that some students are looking for permission to experiment with these tools. There are a lot of mixed messages about AI coming at them from every direction. By providing a discipline-inclusive, library-led framework, we offer a neutral ground where students can experiment and learn without the fear of a misconduct charge. This environment allows for deeper questions about the social and intellectual impacts of these technologies.