Two Perspectives on Teaching AI Literacy and the ACRL Information Literacy Framework

On 20 May 2025, in partnership with Choice and the Association of College and Research Libraries (ACRL), the MLA and EBSCO cosponsored a webinar, Technology and Evolving Research Practices in the Humanities, to discuss how AI tools are shaping teaching and research in the humanities. In this discussion, two webinar participants continue the conversation about AI literacy and the ACRL Information Literacy Framework. Elizabeth Brookbank is professor and instruction librarian at Western Oregon University and the author, with H. Faye Christenberry, of the MLA Guide to Undergraduate Research in Literature. Ellen C. Carillo is professor of English and writing instructor at the University of Connecticut and the author of the MLA Guide to Digital Literacy. Responses have been condensed and edited.

MLA: What common challenges do students experience with using generative AI (GenAI), and have there been any surprising benefits to their use of it?

EB: When students test GenAI in an open way—for example, in a library session where they use it for certain tasks and then we analyze its results together—they come to understand that GenAI is actually not very good at doing their assignments. If I’m in the room when students realize this then I can show them resources that will really help them, like the library research databases or the writing center.

ECC: Students often don’t differentiate among GenAI tools and will default to using ChatGPT and other common tools without considering the specific strengths of each tool. And they often don’t even think to use some older tools that are better, more reliable, and even more efficient for certain tasks.

MLA: Describe your approach to AI literacy and helping students work through those challenges or identify any opportunities.

EB: My approach to AI is based in information literacy and specifically the ACRL Information Literacy Framework. The way I teach students to understand a GenAI tool like ChatGPT is the same way I teach them to use and understand Google or Wikipedia or a library database. Librarians have accepted that students will sometimes use Google or Wikipedia to do their research, so we talk to them about the biases, inequalities, and power structures baked into those sites—and what voices and perspectives are largely missing from them. My hope is that we can get to this same place with GenAI.

When I do an example prompt using AI, I teach students to approach its answer the same way we would approach Google results or library search results: by asking questions about the author’s credentials and experience, the publisher, the sources of information and their trustworthiness. With ChatGPT, you can’t answer most of those questions, and we talk about why that’s a red flag. When students see that ChatGPT has created something from whole cloth or has given them unreliable information, they realize this tool might actually be costing instead of saving them time. GenAI is here, the tools are easily accessible, and they are widely used and talked about. As educators, we should be clear about their downsides so students can make reasoned decisions. Just because the tools are ubiquitous doesn’t mean we have to use them.

ECC: Recently, I invited our campus librarian into one of my literature classes to model for students what research looks like in the age of AI. Then, I spent time reviewing with students when the librarian brought AI into the research process (early, for generating key terms and initial ideas) as well as when the librarian favored the library databases instead of AI. While I teach lateral reading as a means of verifying AI’s output, I also make sure that students understand that the sources they will find in library databases have already been vetted. If students filter their searches in databases, such as the MLA Bibliography and Academic Search Premier, they can easily and quickly find credible materials and sources, including peer-reviewed sources. As such, students can save some time during the research process by relying on long-established resources like library databases, because the material therein has already been reviewed for credibility. Students tend to assume that GenAI is always more efficient, but that’s not the case. I also talk about the strengths of different GenAI tools, so students can make informed decisions about which AI platforms to integrate into their research and other projects.

MLA: You both refer to the ACRL Information Literacy Framework in your MLA Guides, and its importance was mentioned at the webinar. How can it help promote AI literacy?

EB: The ACRL framework is a great tool for teaching students about GenAI as part of information literacy, and teaching information literacy with the framework in mind helps students develop the critical thinking skills they need in order to make decisions about using GenAI. Here are some ways the frames can be helpful when teaching students about GenAI:

• Authority Is Constructed and Contextual. Students often agree that it is risky to trust the information from ChatGPT because some of it probably came from untrustworthy sources whose authority can’t be verified.
• Information Creation as a Process. We can discuss with students how GenAI is trained and how that process is different from the creation process for other types of information.
• Information Has Value. This frame is incredibly important when it comes to GenAI tools, which are designed to make a profit for the multibillion-dollar companies creating them.
• The remaining frames—Research as Inquiry, Scholarship as Conversation, and Searching as Strategic Exploration—are where I really get up on my soapbox. The Research as Inquiry frame tells us that good research is iterative and part of the learning process. But AI tools enable the outsourcing of complex thinking, exploration, and creativity: don’t let tech companies outsource human learning and conversations!

ECC: As a literature and writing instructor, I find that two principles in the ACRL framework are especially germane: Scholarship as Conversation and Research as Inquiry. Both promote the value of process and deemphasize the value of answers in academic and research writing. When these principles are integrated into a course, students are reminded that as emerging scholars they are involved in an ongoing process that involves thinking alongside other scholars and exploring ideas rather than looking for answers.

Students often get overwhelmed when asked to develop an original research question, particularly if they are accustomed to reporting what sources say rather than engaging in true inquiry. Encouraging students to use GenAI early in the research process to generate keywords, suggest topics, or even begin filling up a blank page supports the principle of Research as Inquiry by helping students ease into their research and feel less overwhelmed. Similarly, some AI programs such as Litmaps can support the Defining Scholarship as Conversation principle by helping students visualize (rather than just recognize) the multiple perspectives of sources and the role each plays in a scholarly conversation. However, students must be reminded of the value of their own voice in scholarly conversations. If students depend too much on GenAI, they end up losing that voice; instead of enhancing this ACRL principle, GenAI undermines it. Developing assignments that allow students to pursue a subject meaningful to them can help mitigate such dependence on AI. If students really care about a subject, they are less likely to outsource to AI their contribution to the scholarly conversation.

MLA: How do your MLA Guides help foster AI literacy, and are you working on any other projects you’d like to share?

EB: Faye Christenberry and I wrote The MLA Guide to Undergraduate Research in Literature with the ACRL Information Literacy Framework in mind. We encourage students to approach information with a critical mindset and to ask questions to evaluate sources. I find that students have internalized the idea that scholarly sources are the only “good,” or trustworthy, sources for college assignments. Conversely, this means they believe all sources they deem as trustworthy are scholarly. Chapter 1 gives a set of steps and questions to clarify what scholarly sources actually are and to differentiate them from other sources, such as the information they’re getting from AI.

The Guide also contains example searches and strategies for searching library databases, which can be used to demonstrate the different results and quality of information students will get from a library database compared with searching using a GenAI tool or a search engine that offers AI summaries as top results. These examples in the Guide can be juxtaposed with the same searches using ChatGPT, which will probably fabricate some results. The Guide, used in cooperation with the ACRL framework, is thus very helpful for teaching research and information literacy skills and introducing students to alternatives to using GenAI for research.

ECC: The third edition of the MLA Guide to Digital Literacy (forthcoming in early 2026) includes a chapter dedicated to GenAI and weaves attention to GenAI tools throughout the preexisting chapters. Students are taught how to engineer prompts to get the output they desire, as well as how to engage with GenAI in informed and responsible ways. For example, students are educated on the environmental impacts of AI, the ethical questions surrounding its use, and the privacy issues AI raises. The Guide also outlines how AI is trained and the limitations and biases that result from this training. Ultimately, the Guide encourages students to explore the many uses of AI while maintaining a healthy skepticism about the idea that GenAI is always better and more efficient than other research tools.

My book The Radical Case for Teaching Skim Reading in First-Year Writing (forthcoming in fall 2025 from Utah State UP), dedicates a chapter to how reading pedagogies in our first-year writing classrooms and the field of writing studies more broadly must incorporate the impact of evolving AI technologies. I explore, for example, how arguments against using AI for reading support are making value-laden assumptions about reading, texts, and the reader-text relationship. These assumptions, and the biases therein, I conclude, will only hurt our students, who will need to read in a range of ways both in school and in their civic lives. AI tools can support students as they read for various purposes across their courses.