AI Holds Promise and Pitfalls for Accessibility, NASCIO Paper Finds

A new paper from NASCIO explores the role of generative AI in improving accessibility for people with disabilities, finding both promising use cases and current limitations that require careful consideration.

Key recommendations for state technology leaders: The paper offers four main suggestions to guide the effective use of generative AI for accessibility:

• Engage all stakeholders, including people with disabilities, when evaluating AI tools to ensure they meet the needs of diverse users.
• Cultivate inclusive data sets to combat potential bias that could perpetuate accessibility issues if AI models are trained on inaccurate or exclusionary data.
• Embrace transparency in AI development to empower end users to hold companies accountable and make informed decisions about the tools they use.
• Confront and navigate the current limitations of AI to select the most effective products and approaches for improving accessibility.

Tangible benefits and future potential: Generative AI has already demonstrated measurable improvements for accessibility in certain use cases:

• Speech-to-text AI models have reduced word error rates by 26% for people with atypical speech patterns due to disabilities.
• AI-powered website decluttering and summarization tools have provided benefits for users with dyslexia.
• As the technology evolves, user personalization could further expand information access for people with disabilities.

Limitations to consider: However, the paper also highlights examples where current AI tools have fallen short in fully meeting accessibility needs:

• An AI summarization tool provided incorrect responses to a user with cognitive disabilities, and its writing style felt “robotic” to message recipients when used by someone with autism.
• Inaccurate and delayed captioning at a conference forced a hard-of-hearing attendee to rely on multiple caption feeds to follow along.

Analyzing deeper: While generative AI holds great promise for revolutionizing accessibility, state leaders must remain cognizant of its limitations and potential pitfalls. By engaging diverse stakeholders, using inclusive data, prioritizing transparency, and proactively addressing shortcomings, governments can harness the power of AI to meaningfully empower people with disabilities. However, a measured, thoughtful approach is crucial to ensure AI tools are truly inclusive and effective rather than introducing new barriers in the rush to adopt cutting-edge technology.