When I was a kid, my mom bought a 1979 Camaro Berlinetta. It was a deep blue color with a 350 cubic inch/ 5.7 liter engine. In 1979, that Camaro had 185 horsepower. The ‘79 Z/28 had 245 horsepower.  In comparison, the 2023 Toyota Camry’s base engine is 2.5 liters with 203 horsepower.  The sporty version of the Camry has a 3.6 liter engine and 301 horsepower. All of these engines are naturally aspirated, meaning they don’t use a turbocharger or supercharger.

In 1969, the legendary Yenko Camaro pushed out 425 horsepower from a 427 cubic inch/ 7 liter engine.  My daily driver, a 2019 Audi RS5, pushes 444 horsepower from a 2.9 Liter engine. The 7 liter engine is naturally aspirated, while the Audi engine has two turbos.

What does any of this have to do with AI or accessibility? Well, it appears the skeptics and Luddites are at it again, arguing that “AI won’t solve accessibility”.  It is sentiments like these which hold accessibility back, keeping it in a constant state of playing catch up as new technological innovations happen. It is sentiments like these which give accessibility professionals the reputation of being wet blankets. 

Going back to my car analogy, I believe that the same type of sentiment led to the “Malaise Era” in the US auto industry.  Faced with new safety and efficiency regulations in the 70s, US car manufacturers failed to innovate in the face of these new challenges. Between 1978 and 1982, the Big Three automakers saw an average reduction in sales of around 30%, while Japanese manufacturers like Honda, Toyota, and Datsun ate their lunches. The share of Japanese cars in U.S. auto purchases rose from 9% in 1976 to 21% in 1980. While it would be misleading to blame the Mailaise Era on just a lack of innovation, it is clear that the differences between the US automakers and Japanese automakers at that time did have a lot to do with a failure to innovate.

Failing to innovate is one thing. But actively discouraging others to innovate is even worse.  Well-regarded accessibility thought leaders arguing things like “AI won’t solve accessibility” have a broader impact on the field. The people who hold such an opinion will eventually either see the error of their ways or will find themselves left behind – and possibly jobless. But the bigger problem is the effect this has on others.  This type of sentiment causes others to also discount and ignore new technologies and innovations in a way that further sets the accessibility industry back.

AI is already having a massive impact on accessibility

Among the most important reasons why saying “AI won’t solve accessibility” is so ridiculous is because it ignores the awesome stuff AI is doing already to improve the lives of people with disabilities. 

Tools like Google Live Transcribe, Otter.ai, Zoom’s auto-captioning, and YouTube’s AI captions offer more accessible communication by providing transcription/ captioning in real-time, improving an ability for people who are Deaf or hard-of-hearing a better ability to participate in the workplace. Is it as good as a human? Not yet, but ask anyone who’s used these tools and they’ll tell you these things have gotten way better in a relatively short amount of time!

Apps like Seeing AI (by Microsoft) and Be My Eyes (AI mode) describe scenes, detect faces, read documents, and recognize objects or currency. This promotes a greater amount of independence for people who are blind or have low vision. In fact, the entire field of computer vision has exploded in ways that are going to revolutionize accessibility. We will eventually see this result in blind people driving.

Projects like SignAll and KinTrans use AI to recognize American Sign Language (ASL) gestures in real time. While many who are Deaf or hard-of-hearing are (rightly) skeptical of such products, the fact that they exist at all – and have a high degree of accuracy – means that this communication gap will eventually be closed.

AI is embedded in apps like Waymo, WHILL’s autonomous wheelchairs, or AI-driven smart canes for safe navigation indoors and outdoors. This will improve independence for people with mobility impairments.

Tools like Tobii Dynavox, Proloquo2Go, and Voiceitt use AI to enhance speech generation or recognize atypical speech patterns.  Speech recognition is something that’s been worked on for a long time and has gotten extremely good. Combined with speech generation, you now have closed another communication gap. 

The state of AI today isn’t the state of AI tomorrow

One mistake people make when talking about AI is conflating generative AI with all other types of AI. While it is true that generative AI tends to make use of a lot of other AI technologies, it is important to understand that there’s way more to AI than ChatGPT. And, even more importantly, it is important to remember that the state of AI today isn’t the state of AI tomorrow.

If your experience with AI is limited to using ChatGPT in the last year, it is easy to both recognize its weaknesses while remaining unaware of its progress. Use ChatGPT for math problems, for example, and you’d be well served to verify whatever results it gives you.  But keep in mind that what it gives today is pretty good compared to what results you would have gotten 18 months ago.  

Similarly, if you use ChatGPT to write code, you may find yourself getting frustrated that it does not use consistent coding patterns, cannot keep track of the overall context of what you’re working on, and may undo massive amounts of prior work. But the fact that it can do this at all is nothing short of amazing.

There are real, important, and impactful shortcomings and problems with AI. There’s no way I’d put my life in the hands of a driverless car today. There’s no way I would trust AI for a medical diagnosis or legal help. While I like creating coding tasks with AI, there’s no way I’d use it to write a software product of any complexity.  But I would in the future.

AI is the future of everything, including accessibility

AI might be a  shiny new toy now, but in the future, it’s infrastructure. Just like electricity, the internet, and internal combustion engines before it, AI will become a foundational layer of modern life. Whether you’re using it directly or it’s humming quietly under the hood of the tools you rely on every day, AI will be everywhere. And accessibility will be improved by it.

This is why it’s not just shortsighted to dismiss AI’s role in accessibility—it’s dangerous. AI will power the next wave of assistive technologies, enable more inclusive design by default, and unlock entirely new forms of interaction. The same way fuel injection replaced carburetors and digital dashboards replaced analog gauges, AI is replacing static, rules-based systems with dynamic, adaptive, and personalized experiences.

Accessibility has long struggled to keep pace with the speed of technological change. But AI offers a rare opportunity for accessibility to leapfrog ahead—if we’re willing to embrace it. The people building tomorrow’s AI models are training them on today’s data. If accessibility advocates aren’t at the table now, AI systems will continue to reflect the same exclusions we’ve been fighting for decades. But if we engage, influence, and innovate, we can finally build tools that respond to human diversity instead of forcing people to conform to narrow norms.

Imagine a world where screen readers no longer struggle with poorly labeled websites, because AI fills in the blanks with semantic and contextual understanding. Imagine a world where an autistic person’s communication app learns their speech patterns and preferences over time, getting better the more they use it. Where augmented reality glasses provide Deaf users with live captioning layered over the real world. Where a wheelchair user’s smart assistant can understand their preferences, navigate their home, and even advocate on their behalf.

These aren’t far-fetched sci-fi dreams. They’re coming—and in many cases, they’re already here.

So let’s be bold. Let’s be curious. Let’s treat AI not as a threat to accessibility, but as a tool we must shape—because the future belongs to those who build it. Stop being afraid of change and, instead, work to make sure change is for the better.