Skip to content →

Episode 21: Screen Reader Insights, Accessibility Bug Bounties & AI Agent Barriers

Summary

The Accessibility Breakdown

In this episode of Accessibility Breakdown, Episode 21: Screen Reader Insights, Accessibility Bug Bounties & AI Agent Barriers, Mark Miller and Justin Stockton explore screen reader preferences, accessibility bug bounties, and the barriers created by blocking personal AI agents. Their discussion of WebAIM’s Screen Reader User Survey #11 examines JAWS’ continued popularity, perceptions of free and low-cost alternatives, and how familiarity and everyday workflows influence assistive technology choices.

The hosts then consider Chris Yoong’s proposal for accessibility bug bounties, weighing the potential for ongoing user feedback against challenges in defining and validating issues. Finally, Taylor Arndt’s article sparks a conversation about Amazon’s AI agent restrictions, the importance of using personally configured tools, and why accessibility perspectives need to be included when companies make policies that affect digital access.

In this episode

Screen Reader User Survey #11 Results

WebAIM

  • The hosts explore WebAIM’s findings on screen reader use, proficiency, and navigation habits, noting the survey’s limitations and potential biases.
  • JAWS usage increased slightly despite pricing concerns, while Narrator remained flat—prompting discussion about familiarity and the difficulty of switching tools.
  • 74.6% of respondents considered free or low-cost screen readers comparable to commercial options; the hosts contrast 61.9% of primary JAWS users with 91% of users of alternatives.
  • The conversation considers how cost, support, extensibility, and established workflows may influence screen reader choices.
  • Mark and Justin call for more firsthand perspectives on Narrator’s progress and the features screen reader users rely on every day.

Why does big tech not run Accessibility bug bounties?

Chris Yoong

  • Chris Yoong’s proposal asks whether the security bug bounty model could help companies uncover accessibility barriers.
  • The hosts see potential for ongoing, real-world user feedback that complements technical conformance audits.
  • A bounty program could help identify new barriers as websites and applications change, rather than relying only on periodic assessments.
  • Implementation challenges include reproducible reports, clear eligibility criteria, and disagreements about whether an issue qualifies for a bounty.
  • Mark and Justin treat the proposal as a starting point for discussion, emphasizing the need for boundaries and a consistent validation process.

If You Block My AI Agent, You Block Me

Taylor Arndt

  • Taylor Arndt’s article highlights how blocking personal AI agents can also block access for people who rely on those tools.
  • The hosts discuss Amazon’s restrictions and question whether requiring a company’s own agent meets users’ individual accessibility needs.
  • Personally configured AI agents are compared with familiar assistive technologies: a built-in alternative may not replace a user’s preferred tool or workflow.
  • The discussion emphasizes involving people with disabilities in policy decisions rather than addressing accessibility consequences afterward.

Taylor’s call to action encourages readers to contact Amazon about accommodating agents used for accessibility, while the hosts consider broader questions of user choice and access.

About the author

Mark Miller

Founder & CEO at

Mark SMiller is a digital accessibility consultant who helps organizations build accessible, inclusive digital experiences. He speaks regularly at industry conferences,, CSUN, CVENt and apperes on many webinars and podace. , He trains teams on practical, sustainable accessibility workflows.

Published in Podcasts The Accessibility Breakdown

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *