At GITEX Nigeria in Abuja this past Monday, the nation reaffirmed its commitment to becoming a leading force in artificial intelligence (AI) on the continent by unveiling a multilingual large language model (LLM) designed to accommodate indigenous languages. This groundbreaking initiative, spearheaded by Bosun Tijani, the Minister of Communications, Innovation, and Digital Economy, represents one of the government’s most ambitious efforts to leverage AI for national advancement.
Nigeria’s LLM will support five indigenous languages and various English dialects, a notable departure from conventional AI models that predominantly operate in English and other widely spoken languages. With a populace exceeding 200 million and an astounding diversity of over 500 local languages, policymakers assert that customising AI to reflect local linguistic and cultural realities is essential and strategically beneficial.
Minister Tijani emphasised the significance of this multilingual model during a panel discussion at GITEX Abuja: “When I came in, I was very optimistic that Nigeria needed its own foundational model. We set out to launch a multilingual large language model for our country, encompassing five languages. We have made considerable progress in data collection.”
The reasoning behind this initiative is clear. Millions speak major languages like Yoruba, Hausa, Igbo, and Tiv, which are largely overlooked in global AI frameworks. By prioritising these languages, Nigeria aims to safeguard cultural heritage while enhancing the usability of AI technologies for its citizens.
The National Information Technology Development Agency (NITDA), through its National Centre for Artificial Intelligence and Robotics (NCAIR), is leading the development of this LLM, in collaboration with local AI startup Awarri and international technology partner DataDotOrg.
Creating a large-scale AI model poses considerable challenges, including the necessity for advanced scientific expertise, high-performance computing resources, and consistent funding. Tijani acknowledged these obstacles, stating, “The reality is that the scientific expertise required and the surrounding geopolitical factors are challenges that leaders like myself are grappling with.”
He highlighted the success of countries like India, which developed local AI models while ensuring community involvement and external collaboration. Tijani suggested that Nigeria may need to adopt a blended approach, integrating proprietary and open-source development strategies.
The multilingual LLM is a key component of Nigeria’s overarching National AI Strategy, which has been developing since 2024. Professor Yinka David-West from the Lagos Business School, who has played a vital role in this strategy, explained that it is built upon five foundational pillars: infrastructure, ecosystem development, adoption, governance, and responsible frameworks.
One of the strategy’s most innovative aspects is the AI Collective, a coalition that brings together academia, civil society, and industry stakeholders. The Lagos Business School focuses on research and education, while the Centre for Justice and Legal Development (CJLD) is dedicated to civil society initiatives. Data Science Nigeria fosters the growth of startups and the development of the technical ecosystem.
“The AI Collective aims to cultivate a flourishing ecosystem,” said David-West. “Recognising that academia, industry, talent, and startups are all integral to this process is essential for our success.”
Education is a cornerstone of the AI strategy. Nigeria’s Three Million Technical Talent program engages many fellows in the AI initiative, with about 7,000 expected to contribute to the LLM project. Thus, the program generates a skilled workforce and provides practical learning opportunities.
In addition to technical training, there are ongoing efforts to educate policymakers and civil servants on the fundamentals of AI. Workshops organised by the Lagos Business School are aimed at government officials, covering essential topics such as intellectual property, ethics, and the application of AI in governance.
“If we fail to educate the public sector, our governance framework will remain limited,” David-West stressed.
The multilingual LLM initiative is supported by US$3.5 million in seed funding from various international and local partners, including the UNDP, UNESCO, Meta, Google, and Microsoft.
However, Minister Tijani has expressed concern over the long-term sustainability of relying on donor funds. “For the government to invest millions of dollars into these initiatives, we must determine where that funding will come from. That’s a challenge we must confront,” he stated.
Ultimately, the success of the multilingual LLM will be assessed based on its social impact. By equipping AI systems to understand and generate content in indigenous languages, the project holds the promise of transforming education, healthcare, and public services. Students in rural areas could learn in their native tongues, while local traders and market women would benefit from accessing digital tools in familiar language contexts.
