Using AI for Accessibility

The rise of generative artificial intelligence (AI)  technologies has introduced new means of making courses accessible. When used carefully and thoughtfully, generative AI tools, such as those offered by SDSU, can be used to create and remediate materials for accessibility. Below are a few examples of how to use AI to create more accessible materials.

 

Alternative Text and Long Image Descriptions
 

One of the primary accessibility uses for AI is the generation of alternative (alt) text and long image descriptions for complex images. Alt text is text that is not visible, but will describe images to visually impaired users through the use of screen readers. Long descriptions provide a detailed explanation of a complex image or graphic that goes beyond what can be conveyed with simple alternative text. 

To use AI tools for these functions, upload your image to an AI tool and ask it to provide an alternative text or long description for the image. Read over results, and ensure all necessary information is covered in the descriptions. If results are consistently missing information, tweak your prompts to ensure all information is covered.

Color Contrast Analysis
 

AI can assist in expediting color contrast analyses. While WebAIM's Color Contrast Checking Tool allows users to perform manual tests, AI may be able identify issues more quickly. To perform tests, simply load your image(s) into an AI tool and ask it to check for color contrast issues.

Readability and Inclusive Langauage
 

AI can check for readability. While not a requirement for accessibility, writing content at a reading level as low as is appropriate benefits people with cognitive impairments, people who do not speak English as a first language, and people who may be distracted while reading. Simply copy your writing into an AI tool, and ask it to check for readability, especially if you are targeting a certain reading level.