Six of Africa's largest mobile network operators — Airtel, Axian Telecom, Ethio Telecom, MTN, Orange, and Vodacom — are collaborating with the GSMA to develop artificial intelligence language models trained on African languages. The initiative, announced at Mobile World Congress (MWC) Kigali in October 2025 and detailed in the GSMA's Mobile Economy Africa 2026 report published on Tuesday, aims to address the exclusion of African language speakers from mainstream AI tools. Known as the G6, the operators are working with researchers, startups, and civil society to overcome challenges in data, compute, talent, and policy. Africa has over 2,000 languages, representing more than 30% of the world's total, yet most large language models are built on English and other high-resource languages, leaving millions unable to interact with AI in their native tongues. The project has already delivered a working open Swahili reasoning model, demonstrated at MWC Barcelona 2026 in partnership with MeetKai Zambia, capable of browsing and translating online content in Swahili. This model is intended as a blueprint for expanding AI access to other African languages. The GSMA report projects that closing the language gap could expand the market for digital services and increase mobile's contribution to Africa's economy from $240 billion in 2025 to $290 billion by 2030. Supporting infrastructure includes Cassava Technologies' AI factory in South Africa, which offers GPU computing power and AI services to developers and operators, using Nvidia's Blueprints framework to automate network management. The MTN Skills Academy, accessible across 16 African markets at no data cost, is training local talent for AI deployment. On the data front, Google's WAXAL project, released in February 2026 after collaboration with Makerere University, the University of Ghana, and Addis Ababa University, provides an open dataset of speech in 27 Sub-Saharan African languages spoken by over 100 million people. In March 2026, the GSMA and Zindi launched the African Trust & Safety LLM Challenge, resulting in a safety benchmark for AI models across more than 2,000 African languages. In April 2026, MTN joined a $45 million funding round for ODC, an AI telecoms startup developing network technology tailored to Africa's diverse equipment vendors and unreliable power supply.
The G6 operators are investing in African-language AI while continuing to rely on foreign-owned infrastructure like Nvidia's Blueprints and Google's WAXAL dataset. MTN promotes local AI training through its zero-rated academy but also backed a startup using proprietary foreign frameworks to manage African networks. The push for homegrown models still depends heavily on external technology platforms and funding flows. This creates a paradox where African AI is built with African data but shaped by foreign tools.
Editorial note: AI-assisted opinion, not established fact. Full disclaimer →
