Rosman's Africa-First AI Vision Reshapes Continent's Tech Investment Landscape

Benjamin Rosman’s entry into artificial intelligence was, by his own admission, almost accidental. An undergraduate course that touched loosely on AI concepts was enough to set the course. “There was a course we had that touched on some ideas, very loosely, in artificial intelligence, and I was really excited about that,” Rosman recalls. His guiding philosophy since then has stayed consistent: pursue the most compelling intellectual challenges alongside the most capable collaborators available.

Rosman now holds a professorship at the School of Computer Science and Applied Mathematics at the University of the Witwatersrand (Wits) in South Africa, where he directs the Robotics, Autonomous Intelligence and Learning (RAIL) Laboratory. In 2024, he expanded his institutional footprint by becoming founding Director of the Machine Intelligence and Neural Discovery (MIND) Institute, an organization dedicated to understanding the fundamental science of intelligence across machines, humans, and animals.

Additional reference context is available at https://www.forbesafrica.com/cover-story/2026/08/10/the-south-african-scientist-championing-ai-thats-africa-led/.

His influence, though, extends well beyond campus. Rosman has consistently worked to shift AI discourse away from the insular world of academic research toward broader societal engagement, with particular emphasis on ensuring Africa plays a central role in shaping the technology’s future. That commitment took concrete form through his co-founding of Lelapa AI, a resource-efficient artificial intelligence system built around small language models specifically designed for African languages, dialects, and local operating contexts. The project reflects his conviction that technical expertise must be distributed across geographies rather than concentrated in traditional innovation hubs.

Meanwhile, his advocacy has taken institutional shape through the Deep Learning Indaba, a summer school he co-founded to strengthen machine learning capabilities across the African continent. What began as a focused regional initiative has grown into a global network, now operating satellite events in approximately 47 countries. The scale of that expansion signals both the demand for distributed AI education and the model’s effectiveness in building technical capacity beyond established centers.

Rosman articulates a clear vision for how AI development should unfold. “It’s really important that AI is not just a conversation that happens elsewhere and is imported here,” he states. The framing directly challenges the default pattern in which technological frameworks developed in wealthy nations are subsequently adapted for use elsewhere. His approach inverts that sequence, positioning African researchers and communities as originators rather than adopters.

He is equally clear that technical development alone cannot determine how AI integrates into society. “Building the technology itself is a very technical thing, but the way it interfaces with society in so many ways depends on society. It shouldn’t depend only on computer scientists,” he explains. AI governance, deployment ethics, and social impact, in his view, require input from constituencies well beyond the engineering discipline.

Rosman was appointed a fellow at CIFAR (Canadian Institute for Advanced Research), joining its Learning in Machines and Brains program. TIME also named him one of its 100 Most Influential People in AI for 2025, an acknowledgment of work that bridges fundamental research with practical applications designed for African contexts.

His trajectory raises a question worth watching: as the Deep Learning Indaba’s network continues to expand and Lelapa AI develops its language model capabilities, whether the distributed model of AI development he champions will attract the institutional investment needed to compete with the concentrated resources of the world’s dominant technology centers.