Generative AI

There are several ways that generative AI tools can be used to overcome obstacles in the classroom or enhance the learning process for students. Some examples include:

• Text-to-Speech and Speech-to-Text
  • AI-powered transcription tools and screen readers can help students with visual impairments, dyslexia, or physical disabilities.
• Real-time Captioning
  • AI-driven captioning can support students who are deaf or hard of hearing during lectures or webinars.
• Alternative Content Formatting and Multimodal Learning
  • Generative AI tools can generate summaries, rephrase complex texts, convert text to images or vice versa; these tools support varied learning styles and allow students to engage with course materials in the way that suits them best. These can be especially helpful for English language learners. 
• Customized Learning Support
  • Generative AI tutors can provide tailored explanations of concepts, produce practice questions, evaluate student responses for understanding, or repeat previously-learned concepts for self-paced learning.

While AI tools can be useful in overcoming barriers to inclusive education, there are also barriers to access to take into consideration.

• Bias In = Bias Out
  • Output from generative AI tools can only be as good as the data that the tool is trained on. GenAI may perpetuate stereotypes or exclude disability-related contexts due to biased training data, which can marginalize disabled users.
  • GenAI sometimes produces incorrect or misleading content, or hallucinations, which can mislead users who rely heavily on AI for learning due to accessibility needs.
• Data Privacy
  • Some assistive technologies may transmit sensitive data to third-party AI tools. Be sure to check the terms of service and data use and preservation policies before inputting sensitive data into AI tools.
• Pay to Play
  • Paid versions of common genAI tools are often more up-to-date and capable of more than the free versions of the same tools, making monetary privilege a factor in tool adoption.
• Interface Accessibility
  • Many AI tools are not fully compatible with screen readers, keyboard navigation, or voice commands—making them difficult for users with visual or motor impairments to use independently.