Understanding AI Literacy

Here we offer a framework that identifies and organizes skills and knowledge to help you and your students independently and thoughtfully navigate the opportunities and challenges of generative AI.

Outcomes for this module

This module introduces a framework that identifies and organizes skills and knowledge that you can use to navigate the challenges and opportunities of generative AI technology in education in more thoughtful ways.

After completing this module, you should be able to:

• Describe the skills and knowledge within each of the four domains of this framework
• Apply the AI literacy framework to a real-life educational context
• Identify and commit to actions that can improve your or students’ AI literacy

How do you make decisions about AI?

Imagine a person you know in the process of applying for a scholarship. They tell you they plan to use generative AI chatbots to assist with parts of their application. Consider the situation for a minute. How do you respond? Why?

When responding to a new situation involving complex technology like this, you may have had an initial emotional response. Perhaps you had some questions and recalled what you already know about the topic. Ultimately, you analyzed the situation based on your understanding to come up with a response.

We aim to illuminate and support that decision-making process by providing you with a framework for making informed and intentional choices about using generative AI in a range of educational contexts.

An AI literacy framework

This generative AI literacy framework identifies four intersecting domains of understanding:

• Functional literacy: How does AI work?
• Ethical literacy: How do we navigate the ethical issues of AI?
• Rhetorical literacy: How do we use natural and AI-generated language to achieve our goals?
• Pedagogical literacy: How do we use AI to enhance teaching and learning?

Each domain has suggested objectives at progressive levels of competency. We intend the domains to overlap and inform each other. Those with little or no prior knowledge can begin by building a foundational awareness of the domain. You can then apply that understanding to gain more complex skills and make better-informed decisions. Finally, you can use your skills to create new knowledge and contribute positively to the academic community.

We found inspiration for this framework in Selber’s 2004 Multiliteracies for a Digital Age and the Framework for the Future proposed by Becker, et. al. in 2024. Those frameworks identify three domains: Functional, Critical, and Rhetorical literacies. However, we focus on the ethical issues of generative AI and thus reframed the critical literacy domain in those frameworks as ethical literacy.

A recent faculty survey (From Promise to Practice: Harnessing Gen AI for Evidence-Based Teaching, 2024) reported that faculty proficient in evidence-based teaching practices are more likely to use AI tools. We therefore added a fourth domain, pedagogical literacy, to address this and the concerns of the Teaching Commons.

Miao and Cukurova’s AI competency framework for teachers published by UNESCO in 2024 further inspired the structure and content of each domain.

Functional AI literacy

This domain encompasses the knowledge of how AI works, skills needed to access and operate common tools, and awareness of the ongoing development of AI technology.

You might begin by accessing and operating common AI tools at an introductory level and developing basic prompting skills. You might also examine how AI technology works by defining common AI terminology and describing the process and role of data in the AI development and training process. You can identify different types of AI tools, key companies or organizations, and their major developments in the broader AI technology landscape. Additionally, you can find support resources and familiarize yourself with key policies at the university or in your unit.

As your understanding deepens, you might critique the capabilities and limitations of various AI technologies, use advanced features or tools, practice more sophisticated prompting to generate better outputs, and stay updated on the latest tools and relevant resources.

With mastery, you might customize AI tools for your needs, develop support resources for your community, or coordinate a shared repository of AI tools.

Example educational application: AI chatbots as a composite voice

Large language models (LLMs) that power generative AI chatbots analyze enormous amounts of data from the web to generate the most probable string of words associated with the words prompted by the user. The results might seem impressive, but AI chatbots typically generate patterns of language most common in their training data. That data tends to represent the language conventions and assumptions of dominant perspectives, including any flaws or biases. For most of us using general AI chatbots and straightforward prompting, AI chatbot outputs generally produce an average composite voice aggregated from the internet, scientific papers, and other web sources.

Consider situations in which a composite response serves a purpose; for example, when brainstorming product ideas to have wide appeal or explaining a complex concept to lay audiences. In contrast, consider times when we do not desire a composite response; for example, when we seek highly specialized content expertise, creative and original solutions, or culturally sensitive responses. Understanding how LLMs function can inform how you use AI tools and establish a foundation for more sophisticated use as you build your competency.

Functional AI literacy resources

Continue improving your AI literacy skills by exploring a few of these resources.

• Understand how AI works:
  • Gen AI, Stanford UIT ​​ 
  • AI Guide, AI Pedagogy Project at Harvard metaLAB 
  • Large language models, explained with a minimum of math and jargon, Timothy B. Lee and Sean Trott 
• Begin using AI tools:
  • AI Playground Quick Start Guide, Stanford UIT
  • AI Tinkery, Stanford Graduate School of Education

Ethical AI literacy

This domain includes understanding ethical issues related to AI and practices for the responsible and ethical use of AI tools.

You might begin by examining accuracy, reliability, and implicit bias issues informed by your functional understanding of generative AI technology. You might continue by critically examining research and guidance related to academic integrity, equity, accessibility, privacy, sustainability, and so on. Then you can formulate your position on these key ethical controversies. Stanford’s values of Integrity, Diversity, Respect, Freedom of Inquiry and Expression, Trust, Honesty, and Fairness outlined in Stanford’s Code of Conduct may act as guiding principles around which you can orient your ethical positions.

As your understanding deepens, consider broader ethical issues beyond the education landscape, such as misinformation in politics or healthcare, military applications of AI, labor practices of AI companies, and so on. You and your students can identify and adopt practices that promote individual ethical behavior and establish structures that promote collective ethical behavior. 

With mastery, you might research the impacts of AI on marginalized communities, co-design more ethical AI tools for education, or join collective efforts to enact ethical policies on AI. 

Example educational application: Copyright and authorship

In 2023, the US Copyright Office denied copyright registration for the images in the comic book “Zarya of the Dawn”, produced with a generative AI platform by the author Kris Kashtanova. However, the Copyright Office granted registration for the text, where Kashtanova attested to sole authorship. 

The US Copyright Office provided further clarity in 2023, stating that “a work containing AI-generated material will also contain sufficient human authorship to support a copyright claim” and “In these cases, copyright will only protect the human-authored aspects of the work, which are ‘independent of’ and do ‘not affect’ the copyright status of the AI-generated material itself.”

In cases in which you may want full copyright protection for a work, such as an academic research paper, you must critically examine the implications of using AI tools. On the one hand, AI tools can greatly aid your work; on the other hand, they may lead to unclarity concerning originality, authorship, and ownership. As AI use becomes more collaborative, the distinctions between AI-generated and human-authored material can become more nuanced. Specific policies and recommended practices can vary depending on the context. For example, individual research journals may have different guidelines for AI use, transparency, attribution, and so on.

Ethical AI literacy resources

• Primer on ethical issues in AI for education: Teaching AI Ethics, Leon Furze
• Critical perspectives from educators:
  • One Useful Thing, Ethan Mollick
  • Reflecting Allowed, Maha Bali
  • Cybernetic Forest, Eryk Salvaggio
  • AI + Education = Simplified, Lance Eaton

Rhetorical AI literacy

This domain addresses the skills and knowledge you need to effectively use language to achieve a goal and to analyze the relationship between human-authored and AI-generated language. Here we define rhetoric broadly, beyond just persuasion or argumentation, as the study of how we use language to explore and express ideas and to achieve goals.

You might begin by examining how the structure of your writing, word choice, context, genre, and so on can influence how your audience receives your ideas, messages, and arguments. Rhetorical literacy informs your ability to prompt AI tools to produce quality outputs. You might begin by experimenting with different prompting strategies, evaluating the tone of AI outputs, and identifying patterns in the rhetorical devices used by AI chatbots. Use these insights to adapt and iterate your prompts based on the AI tool or context to produce better results.

As you deepen your understanding, consider how you collaborate with AI tools. Consider this collaboration as a spectrum along which you might shift the tasks, effort, and load dynamically between you and AI tools depending on the situation and your goals. Use your rhetorical awareness to leverage the strengths of human language, like self-expression and building connections, to mitigate the weaknesses of AI-generated language. Combine your rhetorical awareness with your functional and ethical understanding to critique AI-generated content and its place in society.

With mastery, you might refine your voice as an author, develop more inclusive storytelling mediums, or develop AI uses that champion marginalized languages and perspectives.

Example educational application: Writing-to-learn vs. writing-to-communicate

Although we might think of language as primarily a way to communicate with others, we also use language to think and feel. In educational contexts, writing teachers and others sometimes frame this duality as writing-to-learn vs. writing-to-communicate. This framing comes from Writing Across the Curriculum, a common approach to teaching writing in universities.

Writing-to-learn refers to practices, usually informal, through which the act of writing serves as a way of enhancing learning. Exploratory writing, such as brainstorming or freewriting, provides a typical writing-to-learn activity. Note-taking or problem sets also illustrate examples, as the act of taking notes or writing equations helps you remember them. 

Writing-to-communicate, sometimes called writing-in-the-discipline, usually involves formal writing with the underlying purpose of gaining specialized fluency or demonstrating mastery of specific genres and conventions of a discipline. Examples include an academic essay or code for an app.

Imagine the task of writing a list of possible risks of a proposed solution so that you can practice a risk assessment strategy you’ve learned in a course. We can consider this a writing-to-learn task. Therefore, you might use AI as a reflective partner to ask you clarifying questions (as opposed to having AI generate a list of potential risks). In contrast, when asked to synthesize a familiar risk report for new project stakeholders, a writing-to-communicate task, you might use AI for suggested language that uses less jargon to address your new audience. 

Rhetorical AI literacy resources

• On writing:
  • Teaching Writing, Stanford University
  • Writing Across the Curriculum Clearinghouse, Colorado State University Department of English
• AI and writing: Quick Start Guide to AI and Writing, MLA-CCCC Joint Task Force on Writing and AI

Pedagogical AI literacy

This domain encompasses knowledge of effective teaching and learning practices and the ability to integrate them with AI to support student learning and professional development. 

You can begin by exploring learning theories, course design methods, teaching strategies, or study skills that research and experience have shown to enhance learning and promote inclusion. You can also examine AI-powered educational tools and use-cases, attend AI-related professional development events, and explore various pedagogical perspectives from education experts.

As you deepen your understanding, you might identify specific evidence-based practices applicable to your context that promote critical thinking, problem-solving, well-being, and inclusion. You can then develop practices and structures informed by your functional, ethical, and rhetorical understanding that leverage AI to enhance those practices. At the same time, you can identify effective practices that AI could undermine and adapt your practices and structures to mitigate those risks.

With mastery, you might assess the effects of AI integration in your work to inform further pedagogical innovations, organize a professional learning community to explore AI-supported teaching practices, or advocate for ongoing professional development resources for your department.

Example educational application: Zone of proximal development

In his theory of learning, Vygotsky (1978) introduced the zone of proximal development. Educators and the learning environment create this intermediary zone between what you can do on your own and what you cannot do at all even with help. Learning and development happen in this zone where the learner experiences the task as challenging but not overwhelming or too easy. 

A supportive learning environment can help students develop habits that leverage metacognitive reflection, self-examining what they have learned and how they have learned it to identify when they access their zone and learn deeply. In contrast, learners without the right support might mistakenly believe that the ability to complete a task quickly and easily indicates deep learning. How do your current teaching practices help students learn deeply in their zones? 

You might leverage this insight when designing learning activities for students. Perhaps you ask your students to analyze complex sets of symptoms to make a diagnosis. You designed an activity through which students predict a diagnosis based on a patient case study. However, some students struggle with the current case studies, and searching for more case studies consumes too much of your time. You might use AI to generate many short case studies at varying difficulty levels. Then have students reflect on their learning each time they select a case study to fit what they feel is challenging to them.

Pedagogical AI literacy resources

• Foundational teaching practices:
  • Center for the Integration of Research, Teaching and Learning (CIRTL) @ Stanford, Stanford Center for Teaching and Learning 
  • (Re)Designing Your Course, Stanford Center for Teaching and Learning
  • Inclusive Teaching Guide, Stanford Teaching Commons
• Teaching and learning with AI:
  • ChatGPT Assignments to Use in Your Classroom Today, University of Central Florida
  • AI Hacks for Educators, University of Central Florida 
• Join a community:
  • AI x Education
  • AI in Education Google Group
  • GenAI+Learning@Stanford Slack Workspace, Stanford University

Practice applying the framework

Here we offer a few educational scenarios to help you practice applying the framework. We encourage you to engage thoughtfully with each scenario and discuss them with your colleagues. The following prompts can help you get started.

• What functional, ethical, rhetorical, and pedagogical aspects of this situation do you find most important?
• What do you already know that you can apply to this scenario?
• What further skills, knowledge, or people could help you navigate this scenario?
• How can this scenario inform your work?

Scenario 1

Your family member, an undergraduate student, experiences stress about an essay assignment due in a few days. They tell you “I’m so stressed out. I can’t deal with this essay! Do you think a chatbot would be helpful?” How do you approach this conversation with them?

Scenario 2

You have a colleague working on a research paper they intend to submit for publication. They plan to use generative AI tools to help them with the literature review portion of their work. They ask if you know of any tools that could help them find and organize academic sources but also express concern about any ethical issues that may arise and ask for your advice on how they should proceed. How do you approach this conversation?

Scenario 3

Last year, several students asked about using AI chatbots to help them with homework assignments in a course you teach. You expect more students to have such concerns about AI tools this year. You have a few months before the new academic year to revise your course. How might you approach this situation?

Assess and reinforce your learning

Use this activity to self-assess and reflect on what you learned in this module.

Stanford affiliates

• Go to the Stanford-only version of this activity
• Use your Stanford-provided Google account to respond.
• You have the option of receiving an email summary of your responses.
• After submitting your responses, you will have the option to view the anonymized responses of other Stanford community members by clicking Show previous responses.

Non-Stanford users

• Complete the activity embedded below.
• You have the option of receiving an email summary of your responses.
• Your responses will only be seen by the creators of these modules.

Works cited

• Becker, K. P., Parker, J. L., & Richter, D. (2024, June 18). Framework for the Future: Building AI Literacy in Higher Education. Moxie. https://moxielearn.ai/wp-content/uploads/2024/06/Ai-literacies-white-paper.docx.pdf
• From Promise to Practice: Harnessing Gen AI for Evidence-Based Teaching. (2024, Summer). Top Hat. https://drive.google.com/file/d/1IWA1o0hjluMP1FbaF7ec6Bkh8F07CQBo. 
• Miao, F., & Cukurova, M. (2024). AI Competency Framework For Teachers. UNESCO. https://doi.org/10.54675/ZJTE2084
• Selber, S. (2004). Multiliteracies for a Digital Age. (1st ed.). Carbondale: Southern Illinois University Press. https://muse.jhu.edu/book/38844. 
• U.S. Copyright Office. (2023). Zarya of the Dawn (Registration # VAu001480196). https://www.copyright.gov/docs/zarya-of-the-dawn.pdf. 
• U.S. Copyright Office, Library of Congress.  (2023, March 16). Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence. Federal Register. https://www.federalregister.gov/d/2023-05321. 
• Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, Mass.: Harvard University Press.

Learning together with others can deepen the learning experience. We encourage you to organize your colleagues to complete these modules together or facilitate a workshop using our Do-it-yourself Workshop Kits on AI in education. Consider how you might adapt, remix, or enhance these resources for your needs. 

If you have any questions, contact us at TeachingCommons@stanford.edu. This guide is licensed under Creative Commons BY-NC-SA 4.0 (attribution, non-commercial, share-alike) and should be attributed to Stanford Teaching Commons.