Shamsuddeen Hassan Muhammad

Shamsuddeen Hassan Muhammad is a Senior Lecturer (Assistant Professor) at Bayero University Kano, Nigeria, and a Google DeepMind Academic Fellow and Royal Society Career Development Fellow at Imperial College London. He also holds visiting lecturer positions at the African Institute for Mathematical Sciences (Cameroon and South Africa) and York St John University, United Kingdom. He earned his PhD from the University of Porto and his MSc from the University of Manchester. His research spans computational social science, multilingual natural language processing, and AI alignment, with a focus on low-resource African languages. He has published at top NLP and AI conferences such as ACL, EMNLP, NAACL, ICLR, and NeurIPS, and has won several Best Paper Awards. Dr Muhammad co-founded HausaNLP, a research group that builds NLP datasets and tools for Hausa, one of the most widely spoken languages in Africa. He also founded the Arewa Data Science Academy, a free training initiative that has trained over 2,000 Africans in AI and data science.

AI2050 Project

Large language models are now embedded in everyday life, powering tools people rely on for advice, education, and access to information. Yet the more than one billion people across Africa, speakers of over 2,000 languages, rely on systems built without their voices, values, or languages in mind. These systems apply Western cultural norms as universal defaults, generating responses that can mislead, stereotype, or cause harm in critical areas such as health. Muhammad’s project extends AI alignment to African languages and cultures. Working directly with community members and domain experts, he will teach AI models to reason explicitly about African cultural values before responding, and will release all tools, data, and benchmarks openly for others to build on. The goal is a future where AI safety is defined by the communities AI serves.