By Donna Bungard
Sr. Accessibility Program Manager at Indeed
Billion Strong Global Advisor
With the rapid emergence of artificial intelligence (AI) into various aspects of our society, there has been growing conversation surrounding its implications for the Disability Community. As this field continues to expand exponentially, it becomes increasingly crucial to explore its impact. In the upcoming series of articles, we aim to dive into the multifaceted relationship between AI and individuals with disabilities, covering various aspects, including:
Through these discussions, our objective is to provide readers with a deeper understanding of how the emergence of AI technology intersects with the professional lives of people with disabilities. By shedding light on these important considerations, we hope to foster informed dialogue and facilitate the creation of more inclusive workplaces with greater wellbeing for all.
The internet is filled with speculations about AI; however, rather than delving into that discourse, let’s prioritize discussing strategies for fostering disability-inclusive approaches as this technology evolves. Our focus will be on reimagining team structures, redefining success metrics, and navigating the responsible integration of AI in our hiring practices as they pertain to persons with disabilities.
Technological advancements have historically inadvertently created barriers for underrepresented groups due to a lack of diversity in decision-making spaces. Consider the shortcomings of facial recognition technologies with darker-skinned individuals or voice activation systems that fail to recognize minority accents. Numerous examples illustrate this trend over time.
This raises pertinent questions about AI and its relationship with inclusivity. Despite criticisms regarding AI’s understanding of the trans community or its struggles with body positivity barriers, it tends to perpetuate generalizations surrounding disability as well.
In a recent study, I explored AI’s ability to generate images of persons with disabilities. General prompts predominantly yielded images of individuals in wheelchairs, mostly elderly and Caucasian. More specific prompts, such as ‘a person with autism,’ produced images of young, thin, white men who lacked diversity in expression. Particularly alarming was the difficulty in finding representations of the Black disabled community without explicitly specifying ‘a Black disabled person,’ and even then, the results often perpetuated harmful stereotypes.
These findings underscore a glaring issue: the absence of diverse perspectives in the teams that develop those AI learning models. At Indeed, our AI Ethics team is acutely aware of these risks and actively works to embed equity into our learning models. By leveraging resources such as our internal employee resource groups and collaborating with diverse teams across the organization, we strive to infuse equity into the very DNA of our AI systems.
While this series primarily centers on the intersection of AI and disability, our ultimate goal transcends merely having representation from the disabled community. When we advocate for women, Black individuals, the LGBTQIA+ community, and those with disabilities, we inadvertently reinforce segmented labels that compartmentalize identities. Our world is far more complex and nuanced than these categories suggest; therefore, we must deeply respect the intersections of social identities.
Our definition of success necessitates the recognition of intersectionality into our perspective. This entails considering not only disability but also racial, religious, and gender identities, among others. Just as AI has demonstrated its capability to achieve remarkable feats, it can also be harnessed to navigate the complexities of intersectionality.
Indeed’s Smart Sourcing tool exemplifies this ethos by prioritizing a skills-first approach. By focusing on skills, whether gleaned from resumes or user profiles, the tool facilitates precise matching of qualifications to roles, thereby mitigating bias throughout the hiring process.
Embracing a skills-first approach in hiring offers a clear path forward, enabling us to prioritize candidates’ capabilities and contributions to a role. AI plays a pivotal role in this process by scanning traditional resumes to extract skills, sifting through vast pools of applicants to identify the most aligned skill sets, and presenting results in a streamlined manner for hiring managers to assess with ease. Our Smart Sourcing tool exemplifies this capability and goes beyond conventional methods to optimize hiring outcomes.
“Some 79% said skills-first hiring helped them reduce mis-hiring, and 62% said it helped create a more diverse and inclusive workforce.” – Harvard Business Review
As technology continues to evolve, we remain committed to fostering inclusive opportunities for individuals with disabilities and diverse social identities. Through the conscientious training and maintenance of intelligent systems, we uphold a culture of respect and support for their unique journeys.
Read more about Donna Bungard, the author of this article, by clicking here.
Sign up today
Billion Strong is an identity and empowerment organization designed to bring the billions of voices of persons with disabilities together. An exempt organization described in Section 501(c)(3) of the Internal Revenue Code, registered in the United States, EIN# 86-3015273. Please consult your tax advisor for applicable rules and regulations regarding tax deductions.
Copyright ©2024 Billion Strong. All rights reserved
Designed and developed by RGI for
Billion Strong
