Information Literacy

What is Generative AI?

Generative AI is technology that can create new content such as text, images, or videos. It functions as an advanced tool for content creation across multiple media formats.

Large Language Models (LLMs) are one type of generative AI that specializes in creating text. These systems learn by analyzing millions of examples of human writing, then apply that knowledge to generate new text.

LLMs work similar to an advanced autocomplete feature. They predict what word should come next based on patterns learned during training. This allows them to produce confident-sounding responses, but they do not actually understand topics the way humans do—they excel at predicting what sounds appropriate based on their training data.

Sample Applications of Generative AI:

• Content Creation: Writes articles, generates code, creates music or images
• Accessibility: Makes content more accessible by simplifying complex language, generating alternative text, or presenting material in alternative formats (e.g., creating a podcast from an article).
• Personalization: Tailors content and experiences to individual contexts, preferences, and needs
• Writing Enhancement: Assists in editing and improving existing text for clarity and readability.
• Information Summarization: Processes large amounts of information quickly and extracts key points.
• Concept Explanation: Breaks down complex topics and supports exploration of new ideas.
• Idea Generation: Produces ideas on specified topics, presents different perspectives, and helps organize thoughts into structured formats.
• Scenario Simulation: Can adopt different roles or personas for training, practice, or educational purposes.
• Research Support: Provides topic overviews for research purposes, suggests relevant search terms, may suggest relevant sources.

Limitations and Concerns of Generative AI

• Limited Knowledge Base: Training data does not include the most recent information or specialized content that is not publicly available online.
• Accuracy Concerns: AI systems do not truly understand topics—they recognize patterns. This can lead them to generate incorrect information, fabricate plausible-sounding sources, and provide oversimplified explanations of complex subjects.
• Loss of Human Voice and Perspective: When you use AI to communicate, you lose your own unique voice. Additionally, AI-generated content often strips away attribution, obscuring the original human perspectives and creators from which it learned.
• Inherent Bias: AI models are trained on vast amounts of data from the internet, and they inevitably absorb and may even amplify the societal biases contained within that data.
• Copyright and Privacy Issues: AI development involves using extensive content without explicit permission from original creators. Further, users should exercise caution when sharing personal or confidential information with AI, as it could be used in future training or become exposed.
• Environmental and Ethical Concerns: Training and operating large-scale AI models require immense computational power, leading to significant energy consumption. Furthermore, their development can involve problematic labor practices for data annotation and content moderation.
• Potential Learning Impact: Excessive reliance on AI for writing, summarizing, or problem-solving can hinder the development of crucial skills and the critical thinking that emerges from tackling these challenges independently.

Sources:

Poorvu Center for Teaching and Learning. (n.d.) AI Guidelines. Yale University. https://poorvucenter.yale.edu/ai-guidelines 

Claude AI and Gemini were used to edit this content. The prompts used included:

• "Revise the following content to make sure it is concise, accurate, consistent, and professional."
• "Revise the text to make it easier to understand"
• "Make the tone more professional and less conversational, but keep the simpler language and clear structure"

The final output was lightly edited to ensure that the original ideas behind the content were maintained.

Additional Resources:

Students should treat the use of Artificial Intelligence (AI)-generated material the same as material from other external sources AND adhere to the guidelines for responsible usage of AI outlined below. That is, when an assignment allows for inclusion of material from an external source (e.g., material from published or school-based curricula), students can also consider the use of AI tools, so long as they adhere to the best practice guidelines below, including checks and edits of AI output. Students may also use AI to offer revision suggestions on original work or to break down a task. When in doubt about whether AI usage is permitted for an assignment, students should always consult directly with their professor. 

Regardless of the specific assignment or content, the majority of a submitted assignment should be students’ original work. Quoting directly from or paraphrasing from an AI source or using AI in any other way to complete an assignment without citation is plagiarism and violates Relay’s academic integrity policy. If a student uses AI for a Relay assignment, they must cite their use of the tool (i.e., APA citation). 

Relay Student Guidelines for Responsible AI Usage:

• First, check your employer’s policy on AI. Different schools, districts, and networks have different policies on the use of AI. Before you use any AI tool, be sure you know and follow any applicable policies from your employer.
• Protect student privacy. Never share any personally identifiable information about your students with AI (e.g., student names, email addresses, etc.).
• Protect intellectual property. Many AI tools use information that you put in to train the AI itself (e.g., information you add into the AI can be used in its future responses to anyone). As such, excerpts from sources such as Relay coursework should not be provided to any AI tools.
• Keep the limitations of the specific AI tool in mind. Before you use any AI tool, research its limitations. Common limitations include the use of outdated information in algorithms and the potential for provision of inaccurate information through the tool’s output.
• Edit what you get from an AI tool. Because of the limitations above, you should never take and use output from AI without editing it. 

Perhaps the biggest pitfall to AI is to disregard how an AI tool is designed, it's privacy and ethical implications, and it's accuracy. AI continues to improve in leaps and bounds every day, but its is still error prone. However you use AI you should employ a framework to help you evaluate the tool. One useful framework is the AI Literacy Framework for Learners and Educators from Digital Promise. Despite its focus on education, this framework offers a lens to evaluate an AI tool for any use.

Below is an excerpt of the framework:

Source: Mills, K., Ruiz, P., & Lee, K. (2024, February 21). Revealing an AI Literacy Framework for Learners and Educators. Digital Promise. https://digitalpromise.org/2024/02/21/revealing-an-ai-literacy-framework-for-learners-and-educators/ 

Generative AI tools based on Large Language Models (LLMs) like ChatGPT, Copilot, and Gemini can be useful in helping you identify search terms using concept analysis, determine where to search, and select appropriate filters. For example, you may ask an AI tool: "I'm looking for articles about integrating engineering in secondary science classrooms. What search terms should I use? Where and how should I look for authoritative sources?"

Do note that most general AI tools (e.g., Gemini and ChatGPT) will not have access to libraries like the Relay Library, so the tips will be more generalized. If you need help specifically with the Relay Library, you can chat with a real human librarian 24/7!

AI may also be useful in finding sources in that it may provide information that has citations to sources. However, note that AI tools often hallucinate (make up) information and citations, and may not accurately summarize the information in the citations it provides. It can be helpful as a place to start, but you should always visit the original sources directly and evaluate them to trace the truth of the claims! 

Avoid using generalized generative artificial intelligence (AI) tools based on Large Language Models (LLMs) like ChatGPT as a primary source of information. These tools work similar to an advanced autocomplete feature, relying on predictions and patterns rather than actual knowledge and understanding. Therefore, while they often produce confident-sounding responses, it is common for them to "hallucinate" (make up) information and citations. If you find information from an AI tool, ask it to provide source(s), and visit and then evaluate those sources to track the truth of the claim made by the AI.

Generative AI may also be helpful in summarizing content from sources. However, be aware of intellectual property rights (avoid uploading copyrighted material to AI tools) and know that the summaries may be incorrect or misleading. If possible, it's always best to read the underlying sources yourself, and then use the AI summary to check your understanding.

The following describes how you might evaluate AI for validity, relevance, and bias/perspective:

• Validity: Why can you trust the authority of this source?
  • Generative AI is prone to hallucination (making up information). It is not generally recommended to use AI as a definitive source of information.  Even when AI provides or cites sources, it may not accurately summarize those sources. If you need accurate, valid information, try asking the AI tool for its sources, then visit the source directly.  
• Relevance: Is this source relevant for my information needs?
  • General AI tools like ChatGPT can be helpful for brainstorming, creative writing, etc. They may or may not be helpful for more specific needs such as lesson plan writing; you would need to test the tool out to see if whatever it produces is actually high quality.
  • More specialized AI tools (e.g., MagicSchool AI) may be more relevant if your needs are more specific.
• Bias/Perspective: What perspectives are reflected in the source, and how does that compare/contrast with other perspectives?
  • It is impossible to see the training data that is part of most AI tools to determine what biases/perspectives were present. Normally, generative AI will reflect the most common perspectives and biases given that those are most prevalent in its training data; it also tends to agree with rather than challenge the person using the tool. However, you can prompt it to take a different perspective in order to push your thinking.  Still, it is always helpful to critically evaluate the biases and perspectives present in the output, because those are going to be more obscured because there is no human author/organization behind the information that you can research.

How AI Can Support Writing:

AI can help you:

• Create a project plan for how you will complete a writing project
• Create an outline
• Brainstorm
• Suggest possible introductions and conclusions
• Revise and provide feedback on content: generate alternatives for tricky sentences, identify redundant language/clutter, flag where your argument could be clearer, repeatedly review revised drafts, suggest clearer topic sentences, check for tone consistency, break up long sentences, smooth awkward transitions, check for grammatical errors and typos,
• Format citations
• Translate: consider writing content in your native language or using text to speak, then using GenAI to revise the content into standard academic english

General Tips:

• Consider using specific tools such as Grammarly rather than more generalized AI tools such as ChatGPT or Gemini. Specific tools are more likely to give you accurate advice.
• Feed the AI your ideas, voice, writing patterns - don't let it speak for you. 
• AI shouldn't replace your own original thinking, but should support it. 
• Don't ask AI for facts you can't easily check; AI often hallucineates information. 
• Use specific prompts with AI, not just "make it better"
• Engage in multiple iterations with AI; consider using follow up prompts

Sample AI Prompts to Support Writing:

• "Review my outline and give me feedback on the structure of my essay. Is my logic easy to follow? Do my ideas flow logically from one to the next? Where can I improve the structure to strengthen my argument or make the sequence clearer?"
• "Which parts of my argument are solid, and where do I have logical inconsistencies? What counterarguments to my thesis have I not anticipated?"
• "I'd like you to look at the following text I wrote and edit it to make it sound more natural to a native English speaker. Do only minimal/minor edits without changing the tone of the text, which should remain academic and professional." (Source)
• "Review my essay according to the attached rubric."
• "Act as an expert writing tutor. I need to write an analysis of how technology has impacted higher education. Ask me questions to help me write something insightful. Ask one question at a time, wait for my response, and ask me the next question. After I have answered your questions, produce a potential structure for an analytical report for me." (Source)
• "For the attached document, suggest potential revisions to improve the organization of information and academic voice." (Source)
• "I need to write a ten page paper on the history of public education in the United States that includes at least five academic sources. This is due in three weeks, and I can only work on this on Saturdays. Help me develop a plan for completing this task in time."

Sources:

• Lande, J. (2025, April 8). Using AI to improve your writing (without losing your voice). Indisputably. http://indisputably.org/2025/04/using-ai-to-improve-your-writing-without-losing-your-voice/
• Liu, D. (2023, April 28). Prompt engineering for students -- making generative AI work for you. Teaching@Sydney. https://educational-innovation.sydney.edu.au/teaching@sydney/prompt-engineering-for-students-making-generative-ai-work-for-you/

Additional Resources:

APA Citation (From Relay's APA Citation Guide)

General Guidance

The APA gives the following guidance for citing AI tools: “Unfortunately, the results of a ChatGPT “chat” are not retrievable by other readers... Quoting ChatGPT’s text from a chat session is therefore more like sharing an algorithm’s output; thus, credit the author of the algorithm with a reference list entry and the corresponding in-text citation. You may also put the full text of long responses from ChatGPT in an appendix of your paper or in online supplemental materials, so readers have access to the exact text that was generated.” You may also want to include the prompts you used to generate the responses in your appendix.

Citation Format

Author of the AI tool. (Year of the version used). Title of AI tool (Version of AI tool) [Type of AI tool]. URL of AI tool.

Example: OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat

AI Disclosures

AI can be used for a variety of reasons, and it is not always straightforward when or how you should disclose its use. That said, it is generally becoming best practice to disclose the use of generative AI in your work if it goes beyond a spelling or grammar check.  For example, the APA Journals has a policy that when a generative AI is used in drafting a manuscript for an APA publication, authors must specify how, when, and to what extent AI was used.  

The following recommendation on how to create AI disclosures was modified from resources from Newman University and Monash University:

AI Acknowledgements should include:

• The name of the tool used and a link to the website for the tool
•  A description of the type of task you used the tool to accomplish (e.g. generate text, generate images, edit text, generate code, etc.)
•  A list of the specific prompt(s) used
•  An explanation of how you used the output in your final work

Recommended Format:

I acknowledge the use of [insert AI system(s) and link] to [specific use of generative artificial intelligence]. The prompts used include [list of prompts]. The output from these prompts was used to [explain use].

Examples:

Generated Text

I acknowledge the use of ChatGPT 3.5, July 20 version (https://chat.openai.com/) to generate a case study about the challenges of sustainability and labor ethics in the chocolate industry. I entered the following prompt on August 1, 2023:

• "Create a case study over the operations of a chocolate company detailing the challenges of managing chocolate supply chains in both a sustainable and ethical manner. Identify key challenges. Propose potential solutions, but do not commit to any single approach. Conclude by with questions for students to consider about how they would proceed." 

The output from these prompts was used as the case study for the assignment which was then analyzed in the remaining portion of the assignment.

Generated Image

I acknowledge the use of Bing Image Creator powered by DALL-E (https://www.bing.com/images/create) to generate an image of an alien landscape. I entered the following prompt on August 2, 2023:

• "Alien landscape with multiple suns on the horizon. Digital art."

The output from these prompts was used as a decorative image on my PowerPoint presentation.

Editing/Refining Text

I acknowledge the use of Claude (https://claude.ai/) to improve the organization and academic tone of my essay. I uploaded the text of my essay draft through Claude's attach a file feature and I entered the following prompts on August 3, 2023:

• Original prompt: "For the attached document, suggest potential revisions to improve the organization of information and academic voice."
• Follow-up prompt: "Please provide specific examples of corrections or improvements."

The output from these prompts was used to edit my draft essay. Some specific suggestions for rephrasing were adopted, but the output primarily was used to identify areas that could be strengthened from which I made original edits.

Credits: 

https://newmanu.libguides.com/ai-disclosure