10/31/2024

Members of AFB’s Public Policy team joined leaders in the technology and workforce space to share how AI could affect workers with disabilities at the first AI Skilling Fair held by the AI Workforce and Skills Roundtable and House Taskforce on AI. AFB’s Senior Advisor of Public Policy & Research, Sarah Malaier, spoke during the briefing and shared her thoughts after the event.

New technologies like Artificial Intelligence (AI) is exciting when they open up greater opportunities and make accessible work, education, and community living experiences available to all. Yet, we are well aware that people with disabilities are often overlooked in the development, deployment, oversight, and regulation of new technologies, at least at first. The question is: Will AI create opportunities, or will it further exclude people with disabilities?

People who are blind or have low vision participate in the labor force at a rate 30 percentage points lower than people without disabilities. That means we need to be sure that the use of AI does not create new barriers to getting a job, keeping one, or advancing in one’s career.

Some of the AI-enabled tools that are already available are beginning to show promise for access and accessibility. For example, some tools are transforming access to inaccessible images, documents, and videos. AI enables smartphone users to identify elements of unfamiliar environments without vision, and it is expanding access to captioning and automated speech generation. Still, in the workplace, we have seen time and time again that employers are not using accessible tools and that disability is often an afterthought at best and grounds for exclusion at worst.

In the 2021 Workplace Technology Study, AFB found that about half of blind and low-vision employees reported their employer had adopted inaccessible hardware or software. About 1 in 4 reported that online work-related training was not accessible. About 1 in 5 reported that they were reluctant to request accommodations to which they are entitled because of fear of discrimination.

So that leads us to the following questions:

  • Will the software and web platforms that integrate AI be fully accessible to people with disabilities?
  • Will training tools to develop AI skills and digital literacy be usable by all?
  • Will hardware systems, like AVs, be physically accessible and meet the needs for effective communication and assistance?
  • Will decision-making systems account for the diversity of the disability experience? For example, if we use AI systems to recommend performance evaluations or promotions, is there a mechanism for accounting for a loss of time and productivity resulting from using inaccessible software tools and that isn’t based on the employee's skills and commitment to their job?
  • Will AI create new, fair opportunities for people with disabilities? Will it perpetuate existing stereotypes because humans are flawed? Or, could AI create new harms by systematizing bias that previously only arose some of the time?

AFB's research team is working hard to release an expert-consensus study on how AI currently affects people with disabilities. We look forward to using this research to ensure that AI innovation solves problems with inclusion and opportunity in mind. Please sign up for AFB research updates if you would like to receive the research report in early 2025.