Building trust for data sourcing with the disabled community to build robust AI systems

Abstract

This research explores how to build trust between the disabled community and machine learning and AI developers. This is to encourage people with disabilities to contribute their data to build robust models that cater to a greater population. A conflicting reality exists where people with disabilities want to be accommodated in machine learning and AI systems, but they are afraid of sharing their disability status and data with AI developers. The disabled community is subjected to unfair treatment in their daily lives. Many individuals conceal their disability status to avoid being given inferior service or costly biased outcomes in healthcare, insurance premiums, and employment opportunities. When people with disabilities are not included in the data collection process, the results of the machine learning and AI systems will not cater to the needs and preferences of these minority groups. This research will examine the circumstances and privacy protection strategies will people with disabilities garner trust in contributing their data to develop inclusive AI technology solutions. The research will share the opinions of individuals from the communities with sensory disabilities and mobility issues.