digiflowz
The Gist
- Accessibility imperative. Digital accessibility is mandatory, not optional, for modern organizations.
- AI progress. Generative AI holds potential but needs improvement in providing accessible digital experiences.
- Inclusive development. Including individuals with disabilities in AI development is essential for successful accessibility.
Over the last decade, Alison Walden established an extensive track record empowering organizations to enhance the accessibility of their digital experiences. As the senior director of technology for the digital consultancy, Publicis Sapient, her company has worked with many prominent brands including McDonald’s, Marriott and Nissan. And as the head of Publicis Sapient’s Accessibility Centre of Excellence (CoE), she and her team strive to establish standards for accessible digital processes, experiences and capabilities.
How has generative AI been infused into the digital accessibility realm? Thus far, efforts to apply generative AI to digital accessibility challenges have not been too impressive. Why? Walden said it is likely because they’re building on what will become out-of-date interaction models.
“I’m talking about browser adaptations that use generative AI to attempt to fix website accessibility issues, for example by suggesting alternate text for missing images or guessing headings, or even the latest code-completion tools that suggest code snippets using AI,” Walden said. “Today these lack finesse and it’s still necessary to check the output for errors and bad patterns.”
Digital accessibility is serious business for customer experience and marketing teams. The 2023 Gartner Market Guide for Digital Accessibility advises that “Digital accessibility is no longer a choice; it’s a requirement. In addition to training and process updates, software engineering leaders must assess platform and service vendors in a complex and crowded market.”
According to the CDC, one in four adults in America have a disability. That’s 27% of the country’s population.
In a recent interview with CMSWire, Walden provided valuable insights as to what companies should consider when it comes to digital accessibility and generative AI.
Related Article: Why Web Accessibility Is Crucial for Brand Customer Experience
Creating Trustworthy Talkative Tech
In the near future, Walden said we will likely have different ways to navigate online, including through natural language interfaces that will become more accurate and trustworthy than what is currently available.
If you’ve experimented with the latest AI-powered natural language processing tools, such as ChatGPT, you’ve probably experienced the convenience of conducting intricate searches for information or products without the need to navigate through custom (often inaccessible) website interfaces.
“This is a huge benefit for people with disabilities who may find it cumbersome to wade through and validate lists of search results,” Walden said. “AI technology will easily be able to combine information from different sources, like ‘Find me vegan restaurants with female chefs,’ or ‘Find me a brunch spot by a restaurant that mentions having accessible washrooms.’”
However, the problem with today’s tech is the plethora of unreliable results that often appear in search results generated by AI chatbots, requiring the user to double check results or correct the chat interface to arrive at correct information.
“Additionally, once you find a result, you still need to navigate to that website that is probably inaccessible to complete your task,” Walden said. “Imagine a near future in which the search results have become more reliable and in which AI assistants have gained contextual sense and can help us complete tasks even on websites.”
Breaking Generative AI Barriers: Making Context and Accessibility Key Priorities
According to Walden, creating voice assistants that understand context is key. Currently, generative AI developers are improving speech recognition systems for a broader range of speech types, including those with impairments.
“Synthesized speech is also being used to help people who have difficulty speaking communicate more effectively,” Walden said. “Additionally, generative AI can improve text suggestions for people who use augmentative and alternative communication (AAC) devices to communicate.”
Related Article: Driving Customer Engagement With Inclusive Strategies in Digital Marketing
Unlocking Inclusion: Making Digital Accessibility a Priority From the Start
Do you recall the Internet’s grand promise? Walden does. In the 1990s, as information and businesses flocked online, individuals with disabilities anticipated newfound access to previously unattainable content and services. Unfortunately, this vision remained unfulfilled due to the pervasive inaccessibility of the majority of websites.
For people with disabilities, Walden said the benefits of generative AI will only be realized if developers are intentional about considering digital accessibility up front when training large language models and creating natural language interfaces.
“Even though we’ve made technology that makes smartphones accessible — they come with a range of built-in accessibility features such as screen readers, magnification, speech-to-text, and text-to-speech, which make it easier for people with a variety of disabilities to use the devices — people make inaccessible websites and apps all the time that don’t work with the accessibility features of these phones,” Walden said. “Some of these will be the same people who make the current and future versions of AI.”
AI Accessibility Unlocked: Including People With Disabilities in Planning Inclusive Experiences
According to Walden, the best way to make sure that experiences are accessible is to include people with disabilities in their planning and execution. Generative AI can be used to improve digital accessibility only if it’s trained to do so.
“Ensuring generative AI will help these users will only happen if people with disabilities and technologists who understand how to make things accessible are at the table from day one. If we do this right, there is massive potential for benefits,” Walden said. “Just like how in 2008, Google Maps revolutionized how we find our way around (with massive benefits for people with disabilities), new ways of finding information and interacting with the world enabled by generative AI will radically increase the independence and improve the lives of people with disabilities.”
Google’s Accessibility Innovations: From Maps to Live Captions
Global Accessibility Awareness Day has been observed on the third Thursday of May every year since 2012 by institutions and technology companies worldwide.
In honor of the day, Google shared its recently expanded collection of AI-enhanced offerings aimed at digital access and inclusion including access to its Accessible Places feature in Maps for everyone. The tool allows users to quickly identify wheelchair-accessible entrances and provides more details about wheelchair-accessible amenities such as seating, parking and restrooms. Through a collaboration with business owners, local guides and the Maps community, Google has gathered wheelchair accessibility information for over 40 million businesses worldwide.
In addition, Google’s Live Caption, a tool that utilizes AI to provide real-time captions for audio on Android devices, is being expanded to reach more users — and this summer, updates will include a captions box optimized for Android tablets and a new Live caption for calls feature will allow users to type responses during calls that can be read aloud to the other caller.
And during the recent Google I/O event, significant updates for the Wear OS platform were announced with two new sound and display modes were launched to enhance watch customization and a forthcoming release, Wear OS 4, is scheduled for later this year, promising an improved text-to-speech feature that delivers faster and more dependable performance.
Three Top Digital Accessibility Game-Changers
Stephanie Alston is CEO of Black Girl Group, an agency that connects companies looking to hire more diverse employees to racially diverse job seekers. In a post to LinkedIn, Alston delved into the myriad of ways generative AI is actively driving accessibility and fostering inclusion for individuals with disabilities including captioning and audio description, text-to-speech and speech-to-text, prosthetics, assistive technology and accessibility testing.
But Alston told CMSWire, there are three applications she finds most useful.
- Text-to-Speech and Speech-to-Text: Generative AI technologies have greatly improved accessibility for individuals with visual impairments or those who have difficulty reading. Text-to-speech models can convert written text into natural-sounding speech, enabling visually impaired individuals to listen to digital content. Conversely, speech-to-text models can transcribe spoken language into written text, aiding individuals with hearing impairments or those who prefer written communication.
- Language Translation and Localization: Generative AI models have advanced language translation and localization capabilities. They can automatically translate written or spoken content from one language to another, facilitating communication across language barriers. This enhances accessibility by allowing people to access information and interact with digital content in their native language, regardless of the language in which it was originally created.
- Adaptive User Interfaces: Generative AI can enhance accessibility through adaptive user interfaces that cater to diverse needs. These interfaces can dynamically adjust their layout, font size, color contrast, or interaction methods based on user preferences or accessibility requirements. This flexibility ensures that individuals with visual, cognitive, or motor impairments can customize their digital experience to suit their specific needs, making technology more inclusive.
“It’s important to note that while generative AI has made significant strides in enhancing accessibility and inclusion, there are still challenges and limitations to overcome,” Alston said. “Continued research, development and collaboration are crucial to further advancing accessibility technologies.”
