Back in May, Brittle Paper covered the Masakhane African Languages Hub’s open call for the LINGUA Africa initiative, a partnership with the Gates Foundation, Microsoft’s AI for Good Lab, and Google.org offering up to $250,000 in grants for AI projects centered on African languages. Nearly three months on, the results are in: from more than 800 applications spanning 64 countries, 85 percent of them from Africa, the Hub has selected 26 awardees, unveiled during Deep Learning Indaba 2026 in Lagos.

The winning projects were chosen across three categories — Data Creation, focused on building and curating datasets and language resources; Model or Tool Development, covering models, benchmarks, and infrastructure; and Sectoral Applications, which deploys language technologies into real-world settings with measurable impact. Support comes through two funding tracks, Cash + Compute or Compute Only, depending on the project’s needs.

Among the standout awardees, Arusha Technical College in Tanzania is building the country’s first substantial dataset for Tanzanian Sign Language, addressing a glaring gap in accessibility-focused AI work. Nigeria’s HausaNLP is tackling something subtler but just as urgent: models that can actually register the differences between Hausa dialects like Kananci, Sakkwatanci, Zazzaganci, and Katsinanci, rather than flattening them into a single standardized version. On the applied side, organizations including the World Health Organization, Uganda’s Infectious Diseases Institute, and Kenya’s Ushahidi are putting language AI to work in public health and emergency response, building tools meant to function across dozens of regional languages where they’re needed most.

Chenai Chair, Director of Masakhane African Languages, framed the results as proof of a growing field, thanking every applicant and reaffirming the Hub’s commitment to treating African languages as central to AI’s future. The full list of 26 awardees is available via the Masakhane African Languages Hub Substack.