Chinasa T. Okolo, Ph.D.

An emerging area of work is focused on evaluating practitioners' needs and practices when engaging in responsible AI practices, providing valuable knowledge to shape AI development. Our work aims to build upon this research and make a novel contribution by focusing on AI4SG practitioners in the Global South to help understand the challenges of integrating XAI in social impact work and investigate AI4SG practices more broadly. Work under review and further work in progress.

As technology and AI in particular shifts to playing a more important role in the work of CHWs, it will be crucial to understand how XAI can be leveraged to deliver high-quality user experiences and optimal patient outcomes. I begun this work by surveying XAI literature over the past decade to identify papers that discuss techniques of AI applied in the development, deployment, and evaluation of AI-enabled technologies in the Global South. We followed up this work with a field study in Uttar Pradesh India using XAI prototypes to understand and characterize important aspects of model decision-making for CHWs. Future work will entail developing XAI methods catered to users with low levels of technology and AI literacy. Work accepted to COMPASS 2022 (Publication) and CSCW 2024, to appear in the Proceedings of the ACM on Human-Computer Interaction (PACM).

As researchers and technology companies rush in to develop AI applications that aid the health of marginalized communities, it is critical to consider the needs and perceptions of the stakeholders involved in integrating these AI applications into essential healthcare services. We piloted a 3-month study, the first of its kind, examining CHW perceptions of AI in rural India. Drawing on data from 21 interviews, we characterize (1) CHWs’ knowledge, perceptions, and understandings of AI; and (2) the benefits and challenges that CHWs anticipate as AI applications are integrated into their workflows, including their opinions on automation of their work, possible misdiagnosis and errors, data access and surveillance issues, security and privacy challenges, and questions concerning trust. Accepted to CHI 2021 (Publication).

In healthcare, the role of artificial intelligence is continually evolving and understanding the challenges its introduction poses on relationships between healthcare providers and patients will require a regulatory and behavioural approach that can provide a guiding base for all users involved. Our work proposes a framework to set guidelines for how AI-enabled technologies should be formally introduced to patients in healthcare settings. The core principles of this framework are set into 5 parts: Acceptability, Comfortability, Informed Consent, Privacy, and Security. Preliminary work accepted to the Data4Good: Designing for Diversity and Development Workshop (AVI), expanded work under restructuring. (Publication).