Making AI Work For Everyone In Kenya’s Innovation Economy

Kenya’s innovation story has always been written from the ground up. When constraints appear, local builders respond with creativity, resilience, and practical problem‑solving. Artificial intelligence (AI) should be no different. But as AI moves from novelty to necessity, the measure of leadership is changing. The next phase will not be defined by who experiments first, but by who succeeds in embedding AI into daily work, across sectors and skill levels. Without that broad uptake, even the most promising innovation risks remaining theoretical rather than transformative.

Yet the latest research shows that diffusion is far from guaranteed. While global generative AI usage continues to rise, the adoption gap between the Global North and Global South is widening at almost twice the rate. Even the United States, despite leading in frontier AI, has fallen behind smaller, highly digitised economies in workforce adoption. It’s clear that access constraints, not a lack of creativity or ambition, pose the greatest threat to equitable AI progress. For Africa, and for Kenya in particular, this risk cannot be ignored.

Microsoft’s Global AI Adoption in 2025: A Widening Digital Divide report shows that, though Kenya’s appetite for innovation remains strong, the underlying systems required to translate that energy into mainstream adoption are underdeveloped. Although startups, government institutions, researchers, and investors are actively exploring AI solutions, from improving healthcare access to optimising agriculture, national adoption rose only marginally to 8 per cent in late 2025.

Access remains the most immediate challenge. Persistent connectivity gaps and the high cost of smartphones continue to restrict everyday use of AI tools. While Kenya has made significant progress in expanding mobile network coverage, only 33.5% of the population currently uses mobile internet.

Skills shortages create a second barrier. While momentum is building, Kenya still requires the specialised talent required to build, integrate, and manage advanced AI systems. Talent emigration further widens the gap, keeping many organisations stuck at the pilot‑stage of experimentation.

Language and localisation gaps compound the challenge. Most large language models leverage English-language training data, excluding many Kenyan languages and limiting the cultural relevance of AI tools in a country with rich linguistic diversity.

Finally, fragmented regulation slows progress. Overlapping mandates across agencies create uncertainty around governance, privacy, and security, fuelling public hesitation and reinforcing fears around job displacement.

For Kenya, a similar path begins with strengthening digital infrastructure and deepening digitisation. Requiring tertiary students to digitise research projects, for instance, would both create valuable datasets and normalise a digital‑first mindset. Extending digitisation across healthcare, finance, and public institutions will improve data quality and enable AI‑powered optimisation at scale. Importantly, this must be supported by clear regulations that promote innovation while guiding responsible use; policy is not a constraint but a catalyst for trust and adoption.

Professional bodies such as the Kenya Private Sector Alliance (KEPSA) also have a critical role to play. Through training, workshops, and sector‑specific guidance, they can demystify AI, correct misconceptions, and help workers understand its benefits. Public and private early adopters already show what is possible: Kenya Red Cross uses a chatbot to extend mental health support in English and Swahili to millions of underserved people in Kenya, and is transforming access to essential medicines by empowering pharmacy SMEs with AI‑powered tools to improve telehealth services and credit access for underserved communities.

Local language relevance is equally essential. South Korea’s adoption surge only accelerated once AI models became highly effective in Korean. Kenya can follow this path by investing in indigenous language AI. Initiatives such as Paza, recently introduced by Microsoft Research, are starting to show how culturally rooted AI tools can expand access and inclusion. Paza was created to build usable speech AI for low-resource languages, beginning with six Kenyan dialects.

This is Kenya’s moment to convert momentum into scale. The country already has the ideas, the entrepreneurs, and the demand. What’s needed now is coordinated action: investing in infrastructure that expands access, building skills that travel beyond pilot projects, developing AI that speaks local languages, and establishing governance that enables trust. Done well, AI can become more than a technology trend; it can be a practical engine for productivity, inclusion, and shared prosperity, woven into how Kenya works, learns, and grows.

Read Also: Data Governance Across Africa Will Define The Continent’s AI Future

Gerald is the Managing Director, Microsoft Africa Transformation Office