AI, Accessibility, and the Evolving Role of Developers
“This article was published on LinkedIn on February 27, 2025. You can see it here.”
From Vibe Coding to Real Inclusion: Accessibility Needs More Than AI
“This article was published on LinkedIn on February 27, 2025. You can see it here.”
Recently, I interviewed for an accessibility consultant/developer position at a major tech company. The experience was eye-opening, not because the accessibility challenges were new to me, but because of how much emphasis was placed on coding over practical problem-solving.
Over the past ten years, I’ve worked across industries like government, fintech, retail, martech, and non-profits, helping teams build better, more accessible products, among other things. Accessibility has always been a mix of technical execution and user experience. But I’ve noticed a shift: Many companies are betting heavily on AI to solve accessibility problems while overlooking the deep expertise that real professionals bring to the table. Just to clarify, I am not one of those experts. There are people out there who are solely dedicated to the accessibility field. People who have worked for 10, 20 or more years fixing enterprise-level web and mobile applications and have seen it all.
The AI Shortcut Problem
AI has become a game-changer for developers. It speeds up workflows, suggests fixes, and can even generate code. But there’s a catch: Many developers, especially those still learning, rely on AI without fully understanding what it produces. This can lead to what some call “vibe coding”: trusting AI suggestions because they “feel right,” not because they are correct.
This shallow understanding can create serious issues:
-
Bad Accessibility Fixes: AI might recommend ARIA labels where they aren’t needed or miss critical features like a “skip to main content” link. Tools like automated alt-text generators (source) struggle to capture the meaning behind an image, making the web harder to navigate for screen reader users.
-
Harder Debugging: If you don’t fully grasp the code AI gives you, you’ll struggle when something breaks. AI sometimes doesn’t explain its logic properly, and it can take you down the rabbit hole really fast.
-
Unmaintainable Code: Quick fixes might work now, but what happens in six months? If no one understands how the codebase works, future updates become a nightmare.
The Risk of Ignoring Experience
Many accessibility experts are out of jobs right now because companies think AI can replace them. This is a mistake. While AI can optimize workflows, it doesn’t have the lived experience of someone who has spent years making products more accessible for real users. Or the ability to manually test for the navigation to make sense to the user when using tabs or navigating over the list of headers.
The company I interviewed with might have been looking for a more code-heavy accessibility specialist. That’s understandable. But the bigger question is: Should companies be prioritizing code alone, or should they be looking for people who can bridge the gap between AI, accessibility, and real-world problem-solving? People who are taking advantage of what AI can offer as a tool, use their intuition and experience, and make websites and apps accessible? I feel that’s probably the right path to success.
The Future of AI and Accessibility
AI is a tool, not a replacement for expertise. It can suggest solutions, but it can’t advocate for users. It can generate alt-text, but it can’t ensure that text is meaningful. It can suggest accessibility fixes, but it can’t tell you if they truly improve usability.
At least… not yet?
For companies looking to improve accessibility, it’s time to rethink hiring strategies. Accessibility isn’t just about technical implementation; it’s about creating products that truly work for everyone. AI can help, but it’s only part of the equation. The real value comes from experience, problem-solving, and understanding the people who actually use these technologies.