I design AI systems for marginalized communities and study how those systems behave, fail, and get misused once deployed. My fieldwork spans seventeen years across Ghana, Tanzania, Botswana, South Africa, Pakistan, Kenya, and the United States, and organizes around three threads. The first studies how AI systems enable new forms of harm against people who are already marginalized: how scammers target Afghan refugees, how non-consensual image abuse operates differently in honor-based cultures than Western platform policies assume, and how frontline workers in government health systems are forced to falsify the data that AI tools depend on. The second designs and evaluates AI systems for health and high-stakes contexts, including speech-based health services for low-literate parents in Pakistan and culturally-grounded empathy benchmarks for health LLMs. The third builds and measures community digital capacity with marginalized populations rather than for them. I am a Clinical Associate Professor at the University of Michigan School of Information, with over $3.5M in research grants as PI or Co-PI from Gates, NSF, NIH, DFID, Google, UNICEF, and USAID.

My research has been published at ACM CHI, CSCW, and the Web Conference, and recognized with a CHI Best Paper Honorable Mention and a CSCW Diversity & Inclusion Award. I hold a PhD in Design Science from the University of Michigan and an MS from the University of Colorado Boulder, where I was a Fulbright Scholar.

Awards: