Artificial intelligence is advancing quickly around the world. But most AI systems are still built primarily for Western languages and markets.
At MWC 2026 in Barcelona, the GSMA introduced the AI Language Models Initiative, a new program aimed at developing locally relevant AI systems built in Africa, by African developers, for African users. The initiative focuses on one major problem: language.
Across Africa, millions of people cannot fully access digital services because much of the internet still operates primarily in English and a handful of global languages. This language barrier contributes significantly to the continent’s mobile internet usage gap.
To address this, the initiative is developing AI models trained on African languages, cultures, and real-world contexts. This initiative is part of the GSMA’s broader goal of bridging the internet usage gap in Africa with the help of $40 smartphones.
First open Swahili reasoning model unveiled
One of the initiative’s biggest announcements at MWC was the launch and live demonstration of the first open Swahili reasoning model. The model was developed in collaboration with MeetKai Zambia and is capable of browsing online content, translating information into Swahili, interpreting queries, and generating contextual responses.
The system is designed to reduce language barriers when accessing digital services, particularly in regions where Swahili is widely spoken.
Swahili is one of Africa’s most widely used languages, spoken by over 100 million people across East and Central Africa, particularly in countries such as Kenya and Tanzania. Interestingly, the initial collaboration comes from Zambia, a country where Swahili is not the dominant language. Still, the development represents a significant step toward broader African language representation in AI systems.
Why language matters for digital inclusion
Language accessibility plays a critical role in internet adoption. Even when users can afford smartphones and have network coverage, they may still struggle to benefit from digital services if the available content isn’t accessible in familiar languages.
Angela Wamola of the GSMA has previously highlighted that limited local-language content contributes directly to Africa’s high mobile internet usage gap. AI language models could help solve this by enabling local-language search, automated translation of online content, voice assistants that understand regional dialects, and AI tools designed for African communities.
In practical terms, this means users could interact with digital platforms in languages they already speak rather than relying on English or French interfaces.
Infrastructure and compute partnerships
The AI Language Models Initiative is also addressing another challenge: computing infrastructure.
Training and running large AI models requires substantial processing power, something that many African research institutions lack. To support development, the initiative has partnered with technology companies like AMD and Cassava Technologies to help provide expanded compute access for researchers building African language AI systems.
Mapping Africa’s AI talent
Alongside the language models themselves, the GSMA initiative is also introducing tools to strengthen Africa’s AI ecosystem. This includes the creation of a continental AI Talent Map, designed to identify researchers, engineers, universities, and AI startups working on language technologies across the continent.
The goal is to build stronger collaboration between institutions and accelerate the development of locally relevant AI solutions.
Affordable smartphones remain critical
Despite the excitement around AI innovation, the GSMA stresses that device affordability remains the foundation of digital inclusion. Without affordable smartphones, even the best AI tools will remain inaccessible to millions. That’s why the AI initiative is closely linked with the Handset Affordability Coalition’s push for $40 smartphones, also announced at MWC 2026.
According to GSMA Director General Vivek Badrinath, AI’s potential in Africa depends on making devices accessible.
“Affordable smartphones are the gateway to digital and financial inclusion, economic opportunity and innovation. 3.1 billion people have mobile coverage but are not connected to the mobile internet. Together with the G6 group of leading African operators, we are sending a clear demand signal to bring low-cost 4G devices to market. In a global context of rising memory costs, governments have an important role in bridging the usage gap. Removing taxes and import duties on entry-level 4G smartphones will be critical to achieving scale.”
Memory-intensive technologies are required for modern AI capabilities, including on-device processing and language models. Rising memory prices, therefore, present a risk both to affordable smartphones and AI development.
Still, the GSMA believes Africa has a unique opportunity. With strong mobile network coverage and a rapidly growing developer community, the continent could become a major hub for AI built around local languages and real-world needs.
If that happens, millions of new users could finally access the digital services that much of the world already takes for granted.
