Purpose
The Nevada State University Library’s AI Literacy Framework is an internal guide for supporting librarians in extending their expertise in information literacy to the evolving landscape of generative AI and its impact on how information is created, accessed, and shared. It outlines an approach to help faculty and students engage with AI technologies thoughtfully, critically, and responsibly within the library’s teaching and learning contexts. With this purpose in mind, AI literacy is defined as the knowledge and skills that enables students to understand how generative AI works, evaluate AI-generated content, while understanding how to use these tools responsibly and effectively in their academic work.
Competencies
1. Understanding Generative AI Technologies Concepts
Students will develop basic foundational knowledge of how generative AI technologies work and the contexts in which they are used.
- Understand, at a conceptual level, how generative AI produces outputs
- Identify common AI terminology and key concepts
- Distinguish between different types of AI systems and their purposes
Example Activity: Prompting Different AI Tools
Assign students to different generative AI tools.
Provide a single, identical prompt and have each student submit it in their assigned tool. Then, pair students with someone who used a different tool and ask them to compare and contrast the outputs, noting differences in tone, accuracy, detail, and perspective. Conclude with a class discussion on why these differences occur, focusing on how generative AI systems are trained and how they produce responses.
Example Activity: AI System Scenarios
Students analyze short, real-world scenarios that describe how different AI systems are used, such as recommending movies, generating images from text, or detecting fraudulent transactions.
For each scenario, students identify the type of AI system being used, explain its primary purpose, and infer what kinds of data it might rely on to function. By working through multiple examples, students build their understanding of key AI concepts while learning to distinguish between different types of AI systems and how they are applied in everyday contexts.
2. Using Generative AI Tools
Students will make informed decisions about when and how to use AI tools to support their learning.
- Determine when AI use is appropriate and when independent work is more effective
- Refine questions and inputs to improve the relevance, clarity, and usefulness of AI-generated information
Example Activity: Generating Research Questions
Students use an AI tool to generate a research question and keywords.
Use those keywords to find resources via the Library Search (University Library Discovery Platform).
Example Activity: Iterative Question Development
Students begin with a broad research question and use an AI tool to generate an initial response.
They then evaluate the usefulness, clarity, and relevance of that response. Based on that evaluation, students revise their question to be more precise, add context, or clarify intent, and submit it again. This process is repeated for 2–3 iterations. Afterward, students reflect on how changes in their questioning affected the quality and usefulness of the output.
3. Critical Thinking and Evaluation
Students will critically assess AI-generated content and the tools that produce it.
- Evaluate AI outputs for inaccuracies, bias, omissions, and hallucinations
- Assess the reliability of AI tools for academic tasks
- Examine the sources, citations, and evidence behind AI-generated information, including gaps such as outdated, limited, or restricted-access research
Example Activity: Source Comparisons
Compare sources generated by an AI tool vs. Library Search.
Discuss any differences, relevancy, and reliability of sources.
Example Activity: Finding the Gaps
Students are provided with an AI-generated response on a complex topic, such as climate change, healthcare, or education policy, and are asked to critically evaluate its completeness and depth.
They analyze the response to identify missing viewpoints or stakeholders, gaps in evidence or context, and any oversimplifications of the issue. Based on their evaluation, students then propose additional information, perspectives, or credible sources that would be needed to strengthen and more fully support the response.
4. Ethical and Responsible Use
Students will use generative AI in ways that align with academic integrity and broader social responsibility.
- Consider ethical implications of using AI in academic work
- Understand data privacy considerations when using AI tools
- Recognize the environmental and sustainability impacts of AI technologies
Example Activity: Environmental Impact
Students read a short excerpt on the environmental impact of AI and discuss what evidence is used to support the conclusion(s).
Example Activity: AI Course Policy
Students review their course AI policy and discuss its implications for their research process/assignment.
Example Activity: Professional Use Case Study
Students are given a case study in which a professional uses information generated by an AI tool as part of their work.
The case includes the original prompt, the AI-generated response, and verified reference information. Students analyze the AI output to determine its strengths and limitations, identifying which elements are accurate, incomplete, or potentially misleading. They then revise or supplement the information to improve its reliability. Finally, students discuss the ethical considerations and potential real-world impacts of using AI-generated information in professional contexts, including both the risks of errors and the importance of responsible verification.
Glossary
Academic Integrity: The practice of honesty, trust, and responsibility in academic work, including proper citation and avoiding plagiarism or misuse of tools like AI. The commitment to honesty, fairness, and ethical conduct in all academic work. In the context of AI, academic integrity requires that students and researchers accurately represent their use of AI tools, follow institutional policies on AI-assisted work, and take responsibility for the accuracy and originality of what they submit or publish.
Algorithm: A set of step-by-step rules or instructions that a computer follows to perform a task or solve a problem. In the context of AI, algorithms process input data and determine how a model learns, makes predictions, or generates responses.
Artificial Intelligence: A broad field of computer science focused on building systems capable of performing tasks that typically require human intelligence, such as understanding language, recognizing patterns, making decisions, and generating content. AI encompasses a wide range of technologies, from rule-based systems to machine learning models, and is increasingly integrated into research tools, library databases, and productivity platforms.
Bias: In AI systems, bias refers to systematic errors or skewed outputs that arise when a model reflects inequities, underrepresentation, or assumptions present in its training data or design. Bias can manifest in AI-generated text, image recognition, and search result rankings, potentially reinforcing harmful stereotypes or marginalizing certain groups.
Chain-of-Thought Prompting: A technique that encourages an AI system to show or follow step-by-step reasoning when generating a response.
Generative AI: A category of artificial intelligence systems designed to create new content such as text, images, audio, or code by learning patterns from large datasets. Tools like ChatGPT, Gemini, and Claude are examples of generative AI.
Hallucinations: A term used to describe instances when an AI system generates information that is factually incorrect, fabricated, or unsupported, while presenting it with apparent confidence. Hallucinations can include invented citations, inaccurate statistics, fictional events, or misattributed quotes. Because AI models generate responses based on statistical patterns rather than verified knowledge, hallucinations can occur even on common topics.
Large Language Model (LLM): A type of AI model trained on vast quantities of text data to understand and generate human language. LLMs, such as GPT-4, Gemini, and Claude, can perform a wide range of language tasks including summarization, question answering, translation, and content generation.
Library Search: Library Search is the library's discovery platform, providing a single search interface for locating books, articles, journals, databases, and other resources available through the library's collections. Unlike general web searches or AI-generated responses, Library Search connects users to curated and institutionally licensed materials. It is the recommended starting point for academic research, offering reliable, citable sources that meet scholarly standards.
Multimodal: Describing an AI system capable of processing and generating more than one type of data, such as text, images, audio, or video, within a single interaction. Multimodal AI tools can, for example, analyze an uploaded image and respond in text, or generate visual content from a written description.
Natural Language Processing (NLP): A subfield of artificial intelligence concerned with enabling computers to understand, interpret, and generate human language. NLP powers technologies such as chatbots, machine translation, sentiment analysis, and voice assistants. Most large language models and AI writing tools rely heavily on NLP techniques to process and respond to user input in conversational language.
Output: The result or response produced by a generative AI system following a user's input or prompt. Outputs can take many forms such as text, images, code, audio, or video depending on the model's capabilities.
Personally Identifiable Information (PII): Any data that can be used to identify a specific individual, including names, email addresses, student ID numbers, dates of birth, and location data.
Prompt: A prompt is the instruction, question, or text that a user provides to a generative AI system to guide its response. The quality and specificity of a prompt directly influences the usefulness of the AI's output.
Training Data: The large collection of text, images, or other content used to teach an AI model to recognize patterns and generate outputs. The quality, diversity, and recency of training data significantly shape what a model knows, how it responds, and what biases it may carry. Most large language models are trained on datasets compiled from the internet, books, and other public sources, meaning their knowledge has a cutoff date and may not include specialized, proprietary, or recently published academic literature.