Most AI literacy frameworks I’ve covered treat the components as roughly equal: modules learners approach in any order. Hervieux and Wheatley’s (2024) Choice White Paper, “Building an AI Literacy Framework,” makes the opposite move. Their six-frame framework is explicitly hierarchical. Foundational understanding has to come first; critique and discourse come later. After spending time with this paper, I think they got the order right.
The empirical anchor is small but pointed. Hervieux and Wheatley (2024) interviewed 15 instruction librarians across Canada and the U.S. in February 2024. 67% had taught AI content in the past year. Most of them didn’t use the ACRL Framework for Information Literacy in Higher Education to build those sessions. That mismatch between widespread AI teaching and the field’s main literacy framework is the structural finding the rest of the paper builds on.
The authors identify four specific gaps: prompt engineering, critical evaluation pushing past authority into ethics and bias, ethical implications including labor and environment, and attribution for AI-edited content. None of these are well-handled by the ACRL Framework.
Why Hierarchical Beats Modular
The most important conceptual move is the shift from modular to hierarchical. The ACRL Framework treats threshold concepts as non-sequential, with learners approaching frames in any order. That works for information literacy because students already understand texts, sources, and authority before they walk into a workshop.
AI literacy is different. Hervieux and Wheatley argue that “although threshold concepts are certainly a core aspect, most learners need a solid foundation for understanding AI before they can apply other literacy frames to the knowledge sets” (p. 12). That’s the core argument. You can’t ask a student to evaluate AI bias if they don’t know what machine learning is. The same logic applies to engaging with discourse: you can’t engage critically without knowing what generative AI actually does.
The framework builds explicitly on Bloom’s taxonomy: Remember, Understand, Apply, Analyze, Evaluate, Create. The six frames map cleanly to those cognitive levels. The order isn’t arbitrary; it matches how learners actually build technical understanding.
The Sixth Frame Is the Most Useful
The novel contribution is the sixth frame: “engage with AI discourse.” Most AI literacy models I’ve seen stop at evaluation. The authors borrow from ACRL’s “Scholarship as Conversation” frame and argue that staying current with rapidly changing tools is itself a literacy competency.
Hervieux and Wheatley emphasize that “to become AI literate, a learner must stay involved in the conversation by engaging with literature, discussion groups, or any other materials that allow them to keep informed on the topic” (p. 14). AI literacy isn’t a one-time achievement; it’s continuous engagement.
Lo’s (2024) survey of academic librarians, which I’ve covered before, hits the same point from a different direction. Lo found that 70% of librarians felt unprepared to adopt AI in the next twelve months. Hervieux and Wheatley’s sixth frame names what closing that gap actually requires: ongoing involvement in the discourse, not a single training session.
What Teachers Should Take
The pedagogical implications transfer beyond libraries. Teachers building AI literacy units in K-12 or higher ed can use the six frames as a scaffold: terminology and conceptual basics first, differentiation between AI types next, then hands-on experimentation, micro-level review, macro-level evaluation, and ongoing engagement with the discourse.
Chiu’s (2025) editorial on AI literacy versus competency, which I’ve covered before, treats literacy as the foundation for competency. Hervieux and Wheatley’s framework gives teachers a concrete sequence for building that foundation.
The 2025 ACRL AI competencies framework, which I’ve covered before, drew on Hervieux and Wheatley’s work. The 2024 paper is the conceptual draft; the 2025 ACRL document is the institutional version.
Where I’d Push Further
The framework’s limits are real, and the authors acknowledge them. The interview sample is small (15 librarians), the average interview ran 30 minutes, and recruitment relied on listservs and conference networks, which probably skewed the sample toward librarians already interested in AI. The framework itself is conceptual, not yet validated through implementation studies.
I’d add one concern. The hierarchical claim is convincing for novice learners but less convincing for those who already have technical foundations. A computer science graduate student doesn’t need to start at Frame 1. The framework would benefit from explicit guidance on entry points for different audiences.
The authors conclude that “while the ACRL Framework for Information Literacy in Higher Education is a good starting point for planning all information literacy interactions, a new framework is necessary to capture the nuance required to teach AI literacy concepts” (p. 15). The order matters. AI literacy isn’t information literacy with a new label; it’s a new pedagogical project that needs its own scaffolding.
Build the foundation first. Critique comes later. Stay in the conversation as the field shifts. That’s the framework worth using.
References
- Association of College and Research Libraries. (2025). AI competencies for academic library workers. American Library Association. https://www.ala.org/sites/default/files/2025-10/acrl_ai_competencies.pdf
- 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.
- Chiu, T. K. F. (2025). AI literacy and competency: definitions, frameworks, development and future research directions. Interactive Learning Environments, 33(5), 3225–3229. https://www.tandfonline.com/doi/full/10.1080/10494820.2025.2514372
- Lo, L. S. (2024). Evaluating AI literacy in academic libraries: A survey study with a focus on U.S. employees. College & Research Libraries, 85(5). https://crl.acrl.org/index.php/crl/article/view/26409/34344