Welcome to the Teaching and Learning Guide on using AI for Accessibility! 

This guide is designed to help you: 

• Explain Universal Design for Learning (UDL) best practices and how AI can assist in implementing those practices

• Recognize the benefits and drawbacks of using AI to make course content more accessible

• Write effective prompts for AI accessibility teaching tasks

Students today represent a range of abilities, learning preferences, and accessibility needs. As an instructor, you know how important it is to meet the needs of your students. Meeting every student’s needs is a tall task, though. AI can help you meet those needs more efficiently and perhaps more effectively if you know its possibilities and limitations. 

When we think about barriers for students with disabilities, we often think about physical spaces. Physical access barriers do limit students’ access to spaces and it’s important to address these barriers (Sanchez-Diaz and Morgado, 2022; Hong, 2015). Instructors should also consider barriers beyond physical spaces when considering how to create a more inclusive environment for all students. 

When it comes to curricular barriers, studies have shown that there is a lack of awareness  among instructors about disability, “and consequently they generally do not know how to implement teaching-learning strategies to satisfy the needs of students with disabilities” (Sanchez-Diaz and Morgado, 2022). 

AI can be a useful first step in improving knowledge about disabilities. Instructors can use AI as a space to ask general or specific questions about how students with disabilities might experience higher education. For example, we asked Claude AI to compile a list of barriers neurodivergent students might experience in higher education and it gave us the following table.

This is just one example of how to use AI to gain more knowledge about the experience of a student with a disability. You could ask AI to write a narrative about a day in the life of a student who is a wheelchair user. Or you could ask AI to explain why a student might need a particular accommodation. You can also use AI to help create more accessible course content. 

We often think of helping students with disabilities in terms of accommodations, which are legally mandated instructional requirements. Accommodations are important for supporting students, but we prefer the Universal Design for Learning (UDL) approach, which encourages instructors to design their courses for all learners from the beginning. This approach minimizes the need to respond to individual accommodations because all learners are already supported in the course design. 

One important aspect of UDL is organizing content in a way that is easily navigable and student-responsive.

Scaffolding

Research on strong instructional scaffolding has shown that scaffolding, “both a noun (the scaffolding as an observable thing) and a verb (scaffolding as an intentional action in the planning process and implementation),” increases the likelihood of student success (TNTP, 2021).

Scaffolding may include any of the following: 

• Using tools such as 

  • graphic organizers

  • sentence frames

  • concept maps

• Teacher support via feedback

• Peer learning such as: 

  • think/pair/share

  • Peer tutoring

  • Peer review

• Exercises like self-evaluation and reflection

Instructional scaffolding provides students temporary supports that are strategically removed over the course of a lesson or unit to foster independent learning. Instructional scaffolds help reduce interference from non-essential content or other cognitive distractions. 

AI can help you scaffold assignments in many ways, including: 

• Creating the scaffolded tools mentioned above

• Helping you formulate useful feedback for students

• Breaking larger assignments into smaller distinct steps

Chunking

Similar to scaffolding, chunking course content can help reduce cognitive load for students. Chunking includes not only text—although breaking long sections of text into sections and subtopics can be helpful—but also other media. You can chunk a video by creating shorter subtopics. A module, activity, or project can be broken into focused tasks or activities with their own instructions and due dates.

When prompting AI to help with scaffolding, consider being clear and intentional about why you need information chunked or condensed. You should also specify that AI should not remove any content when chunking it. For example, you could start with a prompt like “Please help me by chunking the following content to make it easier to navigate for students with disabilities. When chunking content, please use section headings and subheadings and do not remove any words or sentences.” 

Both chunking and scaffolding support students with dyslexia, dysgraphia, ADHD/Executive Function Challenges, and more by preventing cognitive overload and unnecessary stress while increasing a student’s ability to focus on manageable tasks. Chunking can also help visually impaired students better use keyboard navigation and screen readers.

When it comes to delivering content to students, a UDL approach emphasizes designing courses from the start that include captioning and transcription, alt text, and student choice. 

Captioning and Transcription

Captions are not just helpful extra accommodations. They are essential for many students and useful for all students. Captions can benefit students who: 

• Are D/deaf or hearing-impaired

• Study in noisy or quiet environments

• Learn better with text support

• Need help with note-taking

• Are non-native speakers

Many web conferencing tools available at TWU offer real-time captioning in meetings. These include Zoom, Google Meets, and Microsoft Teams. 

Instructors can also use Panopto or YouTube to add and edit captions in a video. At TWU, once a Zoom meeting in Canvas is recorded, it will automatically be sent to Panopto in that course. When using auto-generated captions in Panopto, it’s important to review them for accuracy. Many auto-captioning tools will make mistakes, which significantly impact students who rely on captioning. 

Free tools like Otter.AI and Microsoft Word Online can help transcribe audio and create summaries although they do have monthly upload limits. Otter.AI can also translate English, Spanish, and French. 

Writing Alt Text

Like captioning and transcription, using alt text helps all students, but especially those with vision disabilities. Alt text or alternative text is descriptive text that you add to images. The goal of alt text is to ensure that students who cannot see the image can still understand its content and function. Good alt text will briefly describe the image’s purpose, key details, and any necessary contextual information. 

AI is a useful tool for helping instructors write alt text, but only when we are intentional about AI prompting. If we ask an AI tool like ChatGPT to write alt text, it will often give non-descriptive and nondetailed responses like “graph showing data.”

With more thoughtful prompt engineering, we can get alt text that is more helpful. For example, a good prompt is: 

I have attached an image of a graph to this message. Can you help me write a helpful, engaging, and visually descriptive alt text for this image? The descriptive text should meet WCAG 2.1 AA standards. Please describe the contents of the image so that a blind student using a screen reader will understand the purpose of the image. The description in the alt text should be 120 characters or fewer. If any text appears in the image, please include that in the alt text description.

Even with more thoughtful prompt engineering, AI alt text needs a human editor. AI can help us get started, though. When evaluating alt text, ask yourself, “Would I be able to understand this image if I couldn’t see it?” Often, the understanding starts with the “Why?” Ask yourself, “Why is this image here?”

• Does it support the message?

• Does it add essential information?

• Does it clarify a complex topic?

• Or does it just take up space or detract from the main point?

If you can’t answer the “Why” convincingly, the image might be decorative, or perhaps doesn’t need to be there at all. Asking why ensures your visuals truly serve your content.

For very complex visuals such as charts, diagrams, and infographics, alt text should be a brief summary of the image’s key takeaway. Crucially, the full, detailed description or data should always be provided in the body text, a detailed caption, or an accessible data table near the image. Your alt text may even need to refer listeners elsewhere for the full description, like this: “Bar chart showing increasing student enrollment. See caption below for detailed data.” This approach ensures no students get left out.

Here is a suggested workflow for making accessible alt text: 

1. Start with “Why?”: Justify the image’s purpose first.

2. Write with Purpose (AI optional): If it’s informative, describe the Who, What, and Where.

3. Be Intentionally Decorative: If it’s just for looks, mark it as such.

4. Check at the Destination: Always add or check alt text in the final platform. (Gardner, 2025)

UDL emphasizes providing multiple and flexible ways for students to learn. Student choice is an important practice for accessibility since it allows students to tailor the course to their specific needs (Lowenthal et al., 2020). Offering flexibility up front about how course content is presented and completed also eliminates the need for you to respond to individual accommodation requests. 

Student choice may be most relevant when it comes to submitting assignments.

Alternative Assignments

When creating an assignment, a UDL best practice is to allow students to submit assignments in their choice of format. For example, you may start with a research question or requirement and offer students the choice to submit their assignment as a written essay, oral presentation, multimedia project, artistic expression, or hands-on demonstration. If creating a specific format is not essential to the intended learning outcomes being assessed, then it should not matter which medium a student chooses to represent their learning. It is only essential that all submissions meet all assignment requirements and grading criteria in the scoring rubric. 

Choice boards are a helpful way to visually represent the varying formats students can choose from to submit their work. We’ve included an example of a project choice board below:

AI can be a useful tool for helping you write assignment requirements that include several options. For example, if you have an assignment that you already use that asks students to write an essay, you could ask AI to help you rewrite the assignment guidelines, allowing students to submit the assignment as a presentation or multimedia project. AI could also help you create a project choiceboard.

For more help with using AI to apply UDL best practices, contact your Instructional Designer.

• Delaney, T. A., & Hata, M. (2020). Universal Design for Learning in Assessment: Supporting ELLs with Learning Disabilities. Latin American Journal of Content & Language Integrated Learning, 13(1), 79–91. https://primo-tc-na01.hosted.exlibrisgroup.com/permalink/f/110i679/TN_cdi_doaj_primary_oai_doaj_org_article_bd40d454aaa741ef9a162a3e16136edc

• Gardner, D. (2025, October 7). More than meets the eye: Writing effective alt text. [Workshop notes.] Texas Woman’s University Accessibility Lunch & Learn series, Denton, TX, United States. https://docs.google.com/document/d/1yOIAKLq_AGUPhbL-iH-S9FftphCnezie8eXZf6Vk638/edit?usp=sharing

• Hong, Barbara S. S. (2015). “Qualitative Analysis of the Barriers College Students With Disabilities Experience in Higher Education.” Journal of College Student Development 56, no. 3, 209–26. https://primo-tc-na01.hosted.exlibrisgroup.com/permalink/f/110i679/TN_cdi_proquest_journals_1684200264

• Killam, L. A., Luctkar-Flude, M., Brune, S., & Camargo-Plazas, P. (2022). Redefining Cheating on Written Exams: A Shift Toward Authentic Assessment to Promote Universal Design for Learning in the Context of Critical Caring Pedagogy. Advances in Nursing Science, 45(3), E127–E143. https://primo-tc-na01.hosted.exlibrisgroup.com/permalink/f/110i679/TN_cdi_proquest_miscellaneous_2608453317

• Lowenthal, P. R., Greear, K., Humphrey, M., Lowenthal, A., Conley, Q., Giacumo, L. A., & Dunlap, J. C. (2020). “Creating Accessible and Inclusive Online Learning: Moving beyond Compliance and Broadening the Discussion.” Quarterly Review of Distance Education, 21(2), 1–82. https://primo-tc-na01.hosted.exlibrisgroup.com/permalink/f/110i679/TN_cdi_proquest_journals_2494575422

• Sánchez-Díaz, M. N., & Morgado, B. (2022). Moving toward the Inclusion of University Students with Disabilities: Barriers, Facilitators, and Recommendations Identified by Inclusive Faculty. The Journal of Continuing Higher Education, 70(3), 175–191.https://primo-tc-na01.hosted.exlibrisgroup.com/permalink/f/110i679/TN_cdi_eric_primary_EJ1371061

• TNTP. “Disrupting Barriers to Strong Instructional Scaffolding: A toolkit for Advancing Mindsets, Principles, Practices, and Conditions.” (2021). https://tntp.org/wp-content/uploads/Tools/disrupting-barriers-to-strong-instructional-scaffolding.pdf