Generative AI has the potential to help individuals with disabilities find and excel in jobs. This article discusses those benefits and how companies can build inclusivity into the design of this technology.

While trust is taking top billing in many discussions about generative AI’s design and uses, it’s also important to bring inclusivity and accessibility into the thick of things early on. That means ensuring that individuals with disabilities themselves play an active role in shaping the technology’s evolution, particularly as it pertains to opening job opportunities and carrying out tasks. The benefits of doing so will lift your company and society. In this article, I explain those benefits and offer some essential guidelines to building inclusivity into the design of this technology.

The Potential

Generative AI augments human capability. It can draw from anything that is conveyed through language. Our research suggests that in the coming years up to 40% of the time people spend working could be assisted by AI that understands language and can converse with its users. In my own life, living with multiple sclerosis, I can already see the potential of what such a “copilot” could offer those of us working and living with disabilities — whether finding and securing employment, providing information to help us interpret situations more accurately, or, similarly, offering guidance as we take part in varied tasks that involve language.

An estimated 386 million working age people have some kind of disability; in some countries, despite persistent labor shortages, unemployment has reached 80% among people with disabilities who could otherwise work. Generative AI can help individuals find — and do well in — jobs that they couldn’t previously have held. For example, OurAbility provides an AI-powered platform that utilizes chatbot technology to assist unemployed individuals in finding and securing employment opportunities.

There are many more people with disabilities who are employed than employers know about. Our analysis (published earlier in HBR) showed that 76% of employees with disabilities had not fully disclosed their unique experiences at work (to colleagues, human resources contacts, or supervisors/managers). At the executive level, the number was even greater, with 80% of C-suite leaders and their direct reports concealing disabilities at work. Regardless of whether individuals have disclosed their disabilities to their employers, personalized generative AI-powered services can enhance their contributions to the organization.

The Challenge

The problem is that getting this right isn’t easy. There just isn’t a good track record to build on. We haven’t done well as a society with the digital divide that exacerbates the barriers between persons with disabilities (as well as other marginalized communities) and others. There are one billion of us with disabilities in the world. And right now, less than 3% of the top one million websites in the world offer a fully accessible experience. And when you broaden the lens to consider the rapid proliferation of generative AI tools, the picture doesn’t improve.

There’s also a very real risk that if companies are racing to get “first mover” status in this space, they may overlook the lessons they (hopefully) have learned about accessibility and inclusivity with previous technologies. That could be costly. According to the Centre for Inclusive Design, “the relative cost of retrofitting a product or service to become inclusive will increase significantly over time and can reach up to 10,000 times the cost of making it inclusive by design.” A far better alternative would be to apply the lessons learned so far as broadly and as quickly as possible.

How Companies Can Get This Right

Start with the people who are designing the human-computer interactions that involve generative AI. Improve their knowledge of inclusive, human-centric design principles that take persons with disabilities into account. Remember that disability is highly nuanced and diverse and user research should be conducted with that in mind. By working alongside and collecting feedback from individuals with various types of disabilities and who are neurodivergent, you can ensure optimal, accessible experiences for all.

Then consider the following list of basic requirements for inclusive interfaces:

Keyboard Navigation

Assess all functionalities of the chatbot, including menus, options, and buttons based on WCAG fundamentals for keyboard compatibility. Ask: What design elements would make these accessible to most? What add-ons can be personalized for others? One example of such an add-on is ChatGPT Let’s Talk, a Chrome extension that adds keyboard shortcuts to the interface, which allows users to talk to and hear AI-generated responses in the site.

Alternative Text

Provide contextual descriptions so that visually impaired users who are using an audio interface with a generative AI tool such as ChatGPT can more fully understand the content. For example, Microsoft’s Bot Framework for developers provides guidelines and features that support the inclusion of alternative text.

Voice-Enabled Interface/Speech-to-Text

Integrate voice-enabled interfaces that enable individuals with a broad range of disabilities (e.g., mobility or motor, visual, cognitive, physical disabilities) to interact with generative AI. For instance, Google’s Dialogflow has built-in integration with Google Cloud Speech-to-Text API, allowing developers to create chatbots that support voice-enabled input.

Text/Image-to-Speech

Include accessibility features such as image or text-to-speech technology to support people with dyslexia or vision or motor impairments. One example is Be My Eyes Virtual Volunteer. Powered by OpenAI’s GPT-4 language model, it allows users to send images via the app to an AI-powered Virtual Volunteer, which will answer any question about that image and provide instantaneous visual assistance for a wide variety of tasks.

Color Contrast, Dyslexia-Friendly Fonts, and Clear Language

Use a high-contrast design of the interface so visually impaired users can differentiate among elements. People with dyslexia will benefit from the improved readability of text and the enhanced user experience when you use a plug-in like Dyslexie Font. In addition, it’s important to use clear and concise language in how to use the chatbot and in the chatbot’s responses (in the languages provided) to enable users with cognitive disabilities to understand the conversation easily.

The rewards of designing for accessibility and inclusivity are significant. As an individual living with multiple sclerosis, I have experienced effects that include cognitive difficulties such as trouble with memory and concentration. In such moments, AI-powered tools such as ChatGPT are invaluable to communicate effectively and access information efficiently. It becomes a supportive companion whether I need help organizing my thoughts or retrieving vital details. Its iterative and responsive nature helps me navigate through moments of cognitive uncertainty, or “brain fog,” assisting me to fully engage in conversations and tasks that would otherwise be overwhelming. It helps me maintain my productivity and independence, and ultimately it enhances my overall well-being despite the challenges posed by my condition.

With regular testing and improvement to ensure that your interfaces comply with the latest Web Content Accessibility Guidelines, those benefits will only grow. These guidelines are set, reviewed, and updated by the World Wide Web Consortium, a nonprofit, global, multi-sector community founded in 1994. Interestingly, chatGPT-powered platforms such as Flowy from Equally AI are already on hand to help with that testing. Let’s get to it.