Hi Reader.
Over the past year, I've been experimenting heavily with generative AI, both in my accessibility consulting work and in the projects that fill the spaces between client engagements. Last month, I had the opportunity to share what I've learned during an IAAP webinar on AI for accessibility professionals. (You can watch the session recording here.) One theme surfaced again and again: AI is neither a miracle worker nor an existential threat. It's a tool. A powerful one. But like any tool, its value depends on how thoughtfully we use it.
My most important rule is simple: never outsource your expertise. Use AI to explore, brainstorm, scaffold, and accelerate, but only in areas where you already understand the work well enough to review the results critically. That means:
- Trust expertise, not outputs. If you cannot perform a task yourself, don't rely on AI to do it for you. You can't accurately review what you don't understand.
- Verify everything. AI can be remarkably helpful and remarkably wrong, often in the same response. Confident wording is not evidence of correctness.
- Make it your own. Revise outputs to reflect your judgment, experience, and voice. Generic content is easy to spot and doesn't build trust.
- Prompt deeply, but know when to stop. Detailed conversations with AI often produce better results, but if prompting becomes harder than the task itself, move on.
- Experiment continually. The AI landscape changes weekly. Stay curious, test new tools, and look for practical ways to improve your workflows and expand your impact.
- Stay sharp without it. Regularly complete tasks on your own without AI. Skills that aren't practiced fade, making it harder to recognize mistakes and poor recommendations.
Accessibility professionals have a unique perspective in the AI conversation. We know what happens when technology is deployed without considering real people, real contexts, and real consequences. AI can help us work faster, but it also amplifies inaccessible patterns, spreads misinformation, and erodes skills if we're not intentional.
Several articles in this issue reinforce the same theme. My recent blog post, along with pieces by Anna Cook and Eric Bailey, explores both the promise and the pitfalls of AI in accessibility. Together, they point to a simple conclusion: AI can accelerate our work and expand our reach, but it cannot replace accessible foundations, human expertise, or thoughtful oversight. As AI becomes more capable, accessibility expertise matters more than ever.
Featured content
AI-Generated Code is Inaccessible by Default
AI can generate websites, applications, and code in seconds, but accessibility rarely comes along for the ride. In this article, I explore why AI-generated code is inaccessible by default, how inaccessible patterns in training data continue to shape what AI produces, and why accessibility expertise remains essential even as AI tools become more capable. Drawing on recent industry trends, real-world examples, and my own experience building an AI-generated accessible web app, the post offers practical guidance for developers who want to harness AI's speed without sacrificing inclusion.
Upcoming events
Accessible Marketing for Email, Social Media, and Digital Outreach
Your website may be accessible, but what about the emails, social media posts, and digital content you share every day? Join me on June 11 for a free workshop hosted by Easterseals Massachusetts and MassAbility, where we'll explore practical ways to make your digital outreach more accessible, inclusive, and effective. Whether you work in marketing, communications, fundraising, or community engagement, you'll leave with actionable techniques you can start using right away to reach and connect with a broader audience.
Accessibility around the web
Building a general-purpose accessibility agent—and what we learned in the process - Eric Bailey
GitHub’s experimental AI accessibility agent can review pull requests, detect WCAG issues, and even fix some problems automatically, helping teams catch barriers earlier in development. But the article’s real insight is that the system only succeeds because GitHub is already investing in strong accessibility practices, testing, and human expertise. AI doesn't create accessibility maturity, it amplifies it.
AI Doesn't Fix Accessible Systems. It Depends on Them. - Anna Cook
AI may be moving fast, but this thought-provoking article challenges the idea that it can solve accessibility on its own. When websites and design systems are inaccessible, AI often amplifies those problems rather than fixing them. Drawing on recent accessibility trends and the rise of AI-assisted coding, the author makes a compelling case that accessible foundations matter more than ever. The key takeaway: not every product needs AI, but every product needs accessibility.
Hey Jealousy! - Emily Wojcik
At the recent DISCO Virtually Abled storytelling event, Emily shared a moving and memorable story drawn from her experiences as a wheelchair user. Her writing carries that same honesty and vulnerability. In this essay, she reflects on attending her college reunion and the complicated emotions surrounding identity, belonging, friendship, and aging with a disability. Rather than offering easy answers, she invites readers into the realities of disabled life with humor, insight, and authenticity. If you enjoy thoughtful disability writing grounded in real-world experience, her Substack is well worth a follow.
Have a jaunty June,
|
|
Simon Miner
Pedal Point Solutions |
Is your website accessible?
Need help with web accessibility consulting, testing, treatment, or training? Pedal Point Solutions can help. Start with a free accessibility snapshot of your website.
|
|
Explore Case for Vision
This stand transforms my iPhone into a hands-free magnifier--helping me read, explore, and stay independent wherever I go.
|
You received this email because you either signed up for the Pedalpointer newsletter or requested a free resource. Use the links below to unsubscribe or update your email preferences.