Expanding Equitable Access to Medical AI

Abdullahi's award-winning paper, “AfriMed-QA: A Pan-African, Multi-Specialty Medical Question-Answering Benchmark Dataset,” is a critical examination of how generative AI systems perform when deployed in healthcare settings with limited access to medical professionals.

The research not only evaluates the performance of these systems across diverse geographic and healthcare contexts but also introduces a much-needed benchmark dataset. This resource is designed to promote equitable access to high-quality medical guidance in resource-constrained environments by providing a robust framework for testing and developing new AI tools.

"I feel proud and happy that this work has been recognized," Abdullahi stated. “This project represents more than technical progress; it reflects a commitment to ensuring that AI innovations reach and serve the communities that need them most.”

A Focus on Trustworthy NLP in Low-Resource Settings

Abdullahi's broader research is dedicated to developing reliable and trustworthy Natural Language Processing (NLP) systems. Her work is driven by core challenges, including improving factuality, robustness, and safety, with a particular emphasis on ensuring these qualities hold in high-stakes, low-resource, and multilingual environments.

Foundation in Climate and Public Health AI

Prior to her doctoral studies at Brown, Abdullahi honed her expertise in the intersection of AI and public health. She earned honors and master’s degrees in Computer Science from the University of Cape Town in South Africa.

While there, she collaborated with Professor Geoff Nitschke and researchers at the Council for Scientific and Industrial Research. Together, they developed a crucial project: a climate-informed AI early warning system for disease outbreaks in risk-prone communities, laying the foundation for her current focus on impactful, real-world AI applications.

To learn more you can read Abdullahi’s paper and visit her personal website.