Güd Marketing

AI & ADA Compliance: A Deeper Look!

by Swati Misra on July 25, 2023 3 -minute read

Background

On July 26, 1990, the Americans with Disabilities Act (ADA) was signed into law and helped set the path for a more inclusive environment for all. The introduction of the ADA and the evolution of the World Wide Web led to the development of Web Compliance Accessibility Guidelines (WCAG) — recommended guidelines for making everything on the internet accessible to people with disabilities. Over the years, companies have adopted WCAG practices to provide comprehensive service to all users. Alternative texts, closed captions, color contrast, text sizes and user experience/user interface for accessible web designs are all widely discussed practices. At Gud Marketing, accessibility is a priority, and we strive to provide WCAG 2.1 AA standards to all our users.

Moving Forward with Newer Technologies

The ever-growing field of technology imposes new challenges and opportunities for accessibility. Significant breakthroughs over the last decade came in the form of voice assistants, household robotic machines, self-driving cars and, most recently, generative AI. A common question we hear is, “What, exactly, is AI?” In simple terms, artificial intelligence is a program that imitates human behavior. It was designed to solve complex problems that previously required human inputs and interactions, and it’s been an essential part of computer research about human intelligence integration for years.
AI makes tasks simpler for people with disabilities. Screen readers emerged as an initial and one of the most popular developments in this area. Screen readers like NVDA and VoiceOver are well-trained to identify screen content and read it out to users. Newer systems are being trained to perform even more sophisticated tasks. A speech-to-text conversion tool allows captioning and transcription in real time, making it much easier for deaf people to interact with all video materials online. It can also be integrated with generative AI tools.
Similarly, audio recognition is a game changer for individuals with disabilities. Smart voice recognition home systems like Alexa, Google and Siri have simplified daily tasks and communication. Advanced tools like ProtoSound are designed to identify sounds from the natural environment. Overall, it’s enormous progress that makes life easier for many.

So, Is AI ADA Compliant?

At this point, it’s important to note that AI is a relatively new technology — the algorithms are in the earliest stage of development. It’s premature to judge the efficiency of AI in terms of accessibility and compliance.
For instance, when integrated with a website, an automated testing tool can detect whether an “alt tag” is available. However, it can’t determine whether the language is biased, inaccurate or unclear. This is also the case with HTML and semantic HTML; an integrated screen reader can verify basic HTML codes but not what they do. Semantic HTML gives more clarity on what an object does in a webpage. Hence, new systems should be trained to identify them swiftly.
Another big concern in this area is the ability of technology to identify diversity and develop unbiased outputs. Speech recognition tools are a good example. While they can recognize some diverse accents, they fall short of discerning things like atypical pronunciations, accents from around the globe, speech disabilities and the nuances of ever-evolving inclusive language.
Providing an accessible user interface is also a part of the ongoing conversation. While generative images are possible now, do they lend the most robust user experience to all?

Final Thoughts

There is no "one size fits all" when it comes to accessibility. AI's potential for optimizing web compliance is evident, but it can only be as good as it's designed to be. Only time will tell if the system can imitate diversity — at this point, human judgment remains indispensable.

Swati Misra

Digital Media Specialist
Coffee lover, movie buff, fitness enthusiast, #GoGreen