As the internet becomes increasingly integrated into every aspect of our lives, ensuring that it is accessible to all is more important than ever. AI has emerged as a transformative force in this arena, offering innovative solutions to enhance accessibility on an unprecedented scale. Donald Norman, a renowned American researcher, professor, and author specializing in cognitive science and usability engineering, once said, “The goal is a technology that serves the user, where the technology fits the task, and the complexity is that of the task, not the tool.” This sentiment perfectly captures the essence of what AI can achieve in digital accessibility: technology that simplifies, personalizes, and empowers.

Also Read: Want to Beat FOIA Backlogs? Embrace AI

In the U.S., one in four adults lives with a disability. That’s approximately 93 million people whose ability to interact with the digital world can be hindered by poor website design. As the internet grows—adding around 252,000 new websites daily—the challenge of ensuring accessibility becomes ever more daunting. The implications of inaccessibility are vast, affecting everything from healthcare and education to social connections and daily entertainment. This is where AI steps in, offering the potential to scale solutions in a way that was previously unimaginable.

The AI Impact

According to a recent WebAIM survey, 97% of the homepages for the top 1,000,000 websites found on the internet were unable to be accessed and operated by people with disabilities.

The Web Accessibility Initiative of the World Wide Web Consortium defines digital accessibility through a set of guidelines, referred to as Web Content Accessibility Guidelines (WCAG) as a shorthand. An example of a WCAG guideline is incorporating alternative text (also referred to as ‘alt text’) for any images found on a website for those with various sensory disabilities (including visual, auditory, and tactile requirements). At a high level, these guidelines serve to standardize accessibility practices.

The most effective approach to solving digital accessibility and ensuring websites are following the accessibility recommendations included in WCAG is through a comprehensive approach that combines human-assisted technology, including AI, and developer tools to address digital accessibility from all angles. Human expertise and manual testing ensure accuracy and provide context-specific solutions, while accessibility tools can identify and fix issues at the source. AI automation plays a key role in advancing these efforts at scale by offering the ability to analyze vast amounts of data, identify accessibility issues, and implement enhancements far more quickly than manual processes could ever achieve, allowing for rapid and widespread implementation of basic solutions. However, AI automation only catches about 33% of common accessibility issues – which further highlights the need for a combined approach that includes human-assisted technology and developer tools.

A recent AudioEye study examined the effectiveness and impact of generative AI on identifying, fixing, and communicating accessibility issues that typically require expert review.

The study found that when AI-generated fixes are used as a starting point, human accessibility experts achieved up to a ten times reduction in the time required to identify a code-based solution and communicate how to implement the update to a developer. In addition to this cut in time, the quality of these AI-generated fixes was nearly on par with human experts, particularly in identifying and addressing complex issues. It is certainly promising to see results like these when considering AI’s potential to scale accessibility solutions by enabling experts to address these issues more efficiently and thus provide a more accessible and user-friendly experience in our online world.

AI models are becoming multimodal or able to process and understand multiple types of data, such as images and audio, in addition to text. This process allows our rapidly improving AI models to holistically comprehend torrents of data. Such progress means AI could revolutionize user testing, which has huge implications for digital accessibility. In this industry, AI models could help us better understand how users with disabilities perceive and navigate websites. There may even come a time when AI models can customize a web experience based on the needs of an individual, which would make the internet more personalized and accessible for everyone. In effect, a pervasively accessible internet in which web interfaces adapt to users, regardless of their disability.

Also Read: The Role of AI and Machine Learning in Streaming Technology

Will AI Replace Humans in Digital Accessibility?

Does this mean that AI will replace humans in the digital accessibility industry? In a word, no. These advancements in what AI models are able to achieve will simply act as an assist to the digital accessibility expert, paving the way for more efficient AI-assisted systems and processes.

Early discussions with industry experts, including insights from OpenAI, suggest that while AI models are becoming increasingly sophisticated—processing text, images, and audio in a more holistic manner—human oversight is still essential. AI’s ability to revolutionize user testing, personalize web experiences, and understand diverse user needs is significant, but the nuanced understanding of context and empathy that humans bring cannot be replicated by machines alone.

Circling back to Donald Norman’s quote earlier, the end result of training and progressing AI models, as seen above, is a widespread accessibility technology that serves the user, putting the responsibility of adaptation on the information technology and not the user. This type of advancement would contribute to the ease of access and use by everyone, particularly those with disabilities.

Moving Forward with AI in Digital Accessibility

AI has changed the scope of our experience online with a slew of new and emerging AI-powered models. Those in the digital accessibility space look to AI as the solution to scale their ultimate goal: an accessible online experience for everyone, particularly those with disabilities.

As outlined in AudioEye’s AI study mentioned above, AI was found to speed up the identification of accessibility errors and minor fixes by a factor of ten times. This allows human experts a better starting point in identifying and communicating the correct fix to developers, which is incredibly promising for digital accessibility experts and advocates.

With roughly 93 million people with disabilities in the U.S. attempting to navigate an online world where 97% of homepages are inaccessible to the disabled community, the need for scale is simply the only way forward. Widespread accessibility will only be reached through continued innovation and collaboration, particularly in the implementation and progress of AI in the realm of digital accessibility.

Also Read: AI and Big Data Governance: Challenges and Top Benefits

[To share your insights with us as part of editorial or sponsored content, please write to psen@itechseries.com]