Accessibility Considerations in the Development of an AI Action Plan
This addresses the problem of AI exclusion and risks for disabled people, but it is incremental as it builds on existing accessibility advocacy without introducing new technical methods.
The paper argues for integrating accessibility considerations into AI development to address the needs of disabled individuals, proposing specific domains such as open-source development, accessible interfaces, and mitigating disability-specific biases.
We argue that there is a need for Accessibility to be represented in several important domains: - Capitalize on the new capabilities AI provides - Support for open source development of AI, which can allow disabled and disability focused professionals to contribute, including - Development of Accessibility Apps which help realise the promise of AI in accessibility domains - Open Source Model Development and Validation to ensure that accessibility concerns are addressed in these algorithms - Data Augmentation to include accessibility in data sets used to train models - Accessible Interfaces that allow disabled people to use any AI app, and to validate its outputs - Dedicated Functionality and Libraries that can make it easy to integrate AI support into a variety of settings and apps. - Data security and privacy and privacy risks including data collected by AI based accessibility technologies; and the possibility of disability disclosure. - Disability-specific AI risks and biases including both direct bias (during AI use by the disabled person) and indirect bias (when AI is used by someone else on data relating to a disabled person).