Anyone working in AI literacy right now has plenty of frameworks to choose from. Long and Magerko’s 17 competencies, the ROBOT test, the Hervieux and Wheatley model, the ACRL adaptations, UNESCO’s competency framework. Add Lo’s (2025) latest contribution to that list.
His five-component framework for AI literacy in academic libraries, published in C&RL News, is brief, clean, and useful for library teams looking for a starting structure. The piece is a position essay, not an empirical study, which means we should read it for what it argues, not for what it proves.
Lo (2025) makes the case for libraries as natural leaders in AI literacy. Academic libraries already teach information evaluation, critical thinking, and ethical research. Adding AI literacy is an extension of that work, not a new mission. He puts it directly: “Just as libraries once championed print, digital, and information literacy, we are now well-placed to become key players in advancing AI literacy as technology shapes the future” (p. 120). I agree with the framing. The interesting question is what AI literacy should actually contain.
AI Literacy Guide for Librarians: Lo’s Five-Component Framework
The framework names five components plus three cross-cutting themes. Technical knowledge comes first: a basic understanding of machine learning, algorithms, and neural networks, but not coding-level expertise. Lo wants librarians and learners to know enough about how AI works to explain why a search engine ranks results in a certain way or why a chatbot misreads a question.
Ethical awareness follows. Lo frames this as critical examination of the values and assumptions built into AI systems, with a useful library-specific example: an AI recommendation tool should be examined for fairness, inclusivity, and transparency before it gets deployed.
Critical thinking is the third component, and this is where Lo’s argument lands strongest. Librarians already know how to teach source evaluation. AI just adds a new kind of source. The same questions apply: where does the data come from, whose perspectives shaped the training, what biases might be present.
Practical skills come next. Lo wants hands-on confidence with tools like ChatGPT, Claude, and Midjourney, plus the judgment to know when AI is useful and when human decision-making is the right call.
Societal impact rounds out the framework. Lo flags hiring algorithms that deepen inequality, surveillance that erodes privacy, and AI’s environmental footprint. The component asks learners to think structurally about AI, not just operationally.
The three cross-cutting themes (human-AI collaboration, lifelong learning, equity and access) thread through the whole framework. They’re useful reminders that AI literacy work needs to be inclusive, ongoing, and collaborative.
Lo argues that “as AI increasingly influences our perceptions and decisions, critical thinking empowers people to reclaim their agency and build a more informed, discerning relationship with technology” (p. 121). That’s the right framing. AI literacy isn’t a checklist of facts to memorize. It’s a posture, a way of approaching tools that demand vigilance.
Limitations
Lo’s five components describe what AI literacy should cover, but the paper doesn’t reckon with three issues that newer research has named directly. The first concern is the librarian self-literacy gap. Lo’s framework assumes librarians have the technical knowledge and ethical reasoning skills to teach others. Most don’t yet. Any framework needs to start from frank acknowledgment of the workforce’s current state.
The framework also misses the AI errors angle. Ford and colleagues (2026) made the case that AI literacy works better when it teaches AI’s failures first, with errors treated as features of the system, not isolated bugs.
Lo’s framework lists hallucination concerns under critical thinking but doesn’t make errors central. After the past two years of confident-sounding AI nonsense reaching court filings and parliamentary submissions, errors should be central to the framework, not relegated to a sub-bullet.
A third gap is the vendor-embedded AI question. McCrary (2026) showed that AI is increasingly baked into library databases by default, often without student awareness or institutional review. Lo’s framework treats AI as something users approach as outside tools (ChatGPT, Claude, Midjourney). It doesn’t engage with the AI students encounter passively inside academic library platforms. That’s a major gap given how many research workflows now involve EBSCO Insights, ProQuest Research Assistant, or Scopus AI as default features.
These aren’t fatal critiques. Lo’s piece is a brief position essay, not a comprehensive treatment. But anyone designing AI literacy work in 2026 should pair Lo’s framework with the empirical and pedagogical pieces that have followed it.
Even with the gaps, Lo’s five-component frame has real practical value. Library teams who need an organizing structure for AI literacy planning can use it as a starting checklist. Each component corresponds to learning outcomes that map well onto existing instruction sessions. Technical knowledge fits into a workshop module. Critical thinking extends source evaluation. Ethical awareness grounds the discussion. Practical skills give learners something concrete to walk out with.
The cross-cutting themes also matter. Equity and access especially. AI literacy programs that serve only the most digitally fluent students will widen the gap, not close it. Lo’s framework names this up front, which is the right move.
I’d recommend Lo’s piece to library teams who are starting from zero on AI literacy and need a clean conceptual map. I’d also recommend reading it alongside the Ford et al. (2026) work on teaching AI through errors, the Ali and Richardson (2025) scoping review for the policy context, and the McCrary (2026) piece on student agency in vendor-embedded AI. The four together give a fuller picture than any one alone.
References
- Ali, M. Y., & Richardson, J. (2025). AI literacy guidelines and policies for academic libraries: A scoping review. IFLA Journal, 51(3), 588-599. https://doi.org/10.1177/03400352251321192
- Ford, H., Burrell, A., Monin, M., Narayan, B., & Jethani, S. (2026). Hacking AI chatbots for critical AI literacy in the library. Journal of the Australian Library and Information Association, 75(1), 42-65. https://doi.org/10.1080/24750158.2026.2614000
- Hervieux, S., & Wheatley, A. (2024). Building an AI literacy framework: Perspectives from instruction librarians and current information literacy tools [Choice White Paper]. Choice / Association of College and Research Libraries.
- Lo, L. S. (2025). AI literacy: A guide for academic libraries. College and Research Libraries News, 86(3), 120-122. https://doi.org/10.5860/crln.86.3.120
- McCrary, Q. D. (2026). Are we ghosts in the machine? AI, agency, and the future of libraries. The Journal of Academic Librarianship, 52, 103181. https://doi.org/10.1016/j.acalib.2025.103181
- UNESCO. (2024). AI competency framework for students. United Nations Educational, Scientific and Cultural Organization. https://doi.org/10.54675/JKJB9835