AI for ADA Web Remediation: Is It the Right Choice?
The web, a vast and sprawling network of information, has become a fundamental part of our lives. From education to work, communication to commerce, it shapes how we interact and learn. However, it’s not always as accessible as it should be for everyone. Web accessibility is a pivotal matter that guarantees everyone, including people with disabilities, can perceive, understand, navigate, and interact with the web. But what role does powerful artificial intelligence (AI), like Accessibe, UserWay, and ChatGPT, play in this realm? And, is it suitable for web remediation?
What is AI?
Artificial Intelligence (AI) is a field of computer science that aims to create systems capable of performing tasks that usually require human intelligence. This includes learning, understanding language, recognizing patterns, solving problems, and making decisions.
AI uses various techniques, most commonly machine learning (ML), where machines are programmed to learn and improve from experience. ChatGPT is, of course, a famous example.
ChatGPT
ChatGPT is a state-of-the-art AI language model developed by OpenAI. It leverages machine learning to understand and generate human-like text based on the inputs it receives. ChatGPT and similar machine learning models have vast potential in their applications. Its applications include coding, composing music, writing emails or essays, and playing complex games like Chess.
While some worry that the mass adoption of AI might take away jobs or make it easier for people to cheat or plagiarize, it has made others question if AI could help with ADA web remediation. While AI has shown tremendous potential in various fields, there are reasons why it may not be the best solution for ADA web remediation.
AI and Machine Learning in Web Remediation
AI and machine learning shows great promise in various fields, including web accessibility. Common AI web accessibility solutions include tools like Accessibe, UserWay, and AudioEye. However, even the most sophisticated AI, like ChatGPT, has limitations. Let’s explore how AI and ML can be applied to web accessibility.
AI Testing for ADA Web Compliance
AI systems can be trained to recognize certain patterns, enabling them to evaluate websites against Web Content Accessibility Guidelines (WCAG) criteria. However, this method is not foolproof.
Many WCAG requirements, such as ensuring content makes sense when read out of context or assessing whether a site is easy to navigate, require a nuanced understanding of human interaction that AI currently lacks. In fact, AI can only identify 30% of WCAG 2.1 A and AA success criteria.
David O’Neill, a well-known figure in accessibility and technology, says, “[AI] can identify accessibility issues, but it cannot replace a human being in terms of applying the necessary context or making the relevant decisions.”
AI Remediating Accessibility Issues
When it comes to fixing accessibility issues, AI may fall short. AI can automate some remediation tasks, like adjusting elements on a webpage, such as text size, color contrast, or page navigation. However, this automatic correction can also introduce new accessibility issues. AI algorithms can misinterpret complex web page elements, leading to flawed solutions that don’t fully address accessibility needs or might cause more barriers.
AI Creating Alternative Text Descriptions
AI has made significant strides in image recognition, leading to the idea that it could be used to generate alternative text descriptions for images. However, though AI can identify the objects in a photo, it still struggles to provide meaningful context. For example, an AI might recognize and describe an image as “a person standing in a room,” but a human could provide more context, like “John, laughing while decorating the Christmas tree.”
Predicting and Prioritizing Accessibility Issues
AI and ML algorithms can predict common accessibility issues based on patterns from past data. However, predicting is one thing; prioritizing and rectifying these issues is the real challenge. AI lacks contextual understanding, which is often vital in prioritizing and addressing barriers. For instance, AI might deprioritize fixing small font sizes based on frequency. Still, for an elderly user or someone with visual impairments, this is a critical issue.
So, at a glance, AI offers some promising solutions. But should we entirely trust AI for web remediation?
Limitations of AI for ADA Remediation
While AI and ML have made strides in web accessibility, it is clear they can’t replace the expertise and experience of a web accessibility agency.
AI can certainly help identify some common accessibility issues. But to thoroughly evaluate the accessibility of a website and remediate potential problems, you need a detailed understanding of how people with disabilities interact with digital content. This level of insight requires human empathy, contextual understanding, and years of experience, something AI currently lacks.
AI cannot comprehend the intent of a web page or the nuance behind the design. It can’t understand the user journey in the same way that a professional with experience in web accessibility can.
Manual Auditing: The Gold Standard
While automating the process might be tempting, manual auditing remains the gold standard for ADA web remediation.
A manual audit offers depth and understanding that AI currently cannot match. With a comprehensive understanding of user experience from various perspectives, human auditors can identify barriers to accessibility that AI might miss. This is because professional auditors are experienced in recognizing how individuals with disabilities navigate and interact with web content.
By leveraging the expertise of professionals, you can ensure that every aspect of your website meets WCAG standards. They can understand the unique context and nuances of your website, something which even the most advanced AI, like ChatGPT, cannot fully emulate.
Partner with 216digital for ADA Web Remediation
When it comes to ADA web remediation, consulting professional web accessibility agencies like 216digital provides a distinct advantage. Our team of experts understands the intricacies of web accessibility and how to make your website more accessible and compliant with ADA and WCAG guidelines.
Unlike AI, 216digital offers customized strategies based on a comprehensive understanding of your website and its users. We don’t just look for issues; we also devise the best ways to remediate them in light of your site’s purpose and design.
As David O’Neill points out, “There is a role for automation in accessibility… [But] there is also a clear need for manual testing.” At 216digital, we combine the best of both worlds, using technology to aid in identifying issues while relying on our human expertise for effective remediation.
For comprehensive ADA web remediation, trust the experts at 216digital. Schedule a complimentary ADA Strategy Briefing today.