Artificial intelligence has become one of the biggest policy conversations in Africa over the past few years. Many governments are racing to figure out how to regulate generative AI, manage its risks, and position their economies to benefit from it.
Kenya has joined the list of countries actively trying to regulate it. The East African country has opened its draft National Artificial Intelligence and Emerging Technologies Policy for public participation, offering the clearest picture yet of how one of Africa’s top technology ecosystems plans to approach AI.
However, rather than treating AI as another digital tool that needs rules, the policy draft frames it as a strategic national asset that could determine the country’s economic competitiveness, digital independence, and long-term development.
Importantly, the draft notes that disclosure alone should not replace the right to explanation, human review, or redress, emphasising that simply publishing how an AI system works is insufficient to protect people, because users need clear explanations of particular decisions, the opportunity to check outcomes, and the ability to fix unfair results.
Spearheaded by the Ministry of Information, Communications, and the Digital Economy (MICDE), the policy draft establishes a unified national framework designed to transition Kenya from a technology recipient into a sovereign producer, evaluator, and regulator of intelligent systems. The blueprint offers a compelling playbook for how developing nations can harness frontier technologies without forfeiting their digital sovereignty.
Innovation and national capability
Much of the global conversation around AI has centred on chatbots, copyright disputes, misinformation, and job displacement. While Kenya acknowledges many of these concerns, the draft policy focuses more on infrastructure, compute capacity, data governance, local talent, research, cloud services, and digital sovereignty.
The policy outlines flagship National AI Missions across high-priority strategic sectors, including climate resilience, automated diagnostics for rural healthcare, precision agriculture, and public finance optimisation.
Recognising that sophisticated AI models require accessible infrastructure, the policy introduces the National AI Compute Access Programme, designed to democratise access to compute infrastructure, ensuring that local innovators, researchers, public institutions, and startups are not priced out of the intelligent economy.
To lower the high barrier to entry for local startups and academic researchers, Kenya intends to negotiate tax relief across the East African Community (EAC) to reclassify GPU hardware and AI accelerators as educational equipment.
Victoria Fakiya – Senior Writer
Techpoint Digest
Stop struggling to find your tech career path
Discover in-demand tech skills and build a standout portfolio in this FREE 5-day email course
The country plans to power this compute expansion by leveraging its vast geothermal grid, the sixth-largest globally, positioning it as Africa’s hub for sustainable, green data centres.
Human rights and worker safeguards
Kenya’s framework directly addresses a critical, often-overlooked segment of the AI supply chain: the human workforce.
The country has long served as a key hub for data annotation and content moderation. Hence, the policy establishes binding duty-of-care standards for these workers, mandating transparent pay benchmarks, mental health support, and statutory protections against exposure to harmful content.
Furthermore, the policy embeds a strict human-centred philosophy across all public-sector deployments. Citizens retain the explicit right to know when they are interacting with an automated system, and human-in-the-loop oversight is required for high-risk decisions affecting legal rights, health, or public benefits.
To put these guardrails into motion without stifling innovation, the policy adopts a hybrid governance model. The Steering Committee and National AI Council will provide strategic coordination and maintain a centralised public registry of deployed public-sector AI systems, while a tiered risk framework categorises algorithms by impact, subjecting high-risk deployments to mandatory impact assessments and independent audits.
The Kenya AI Safety Institute (KAISI) will serve as an independent technical centre of excellence responsible for model testing, red-teaming, benchmarking, and incident tracking.
Owning AI
One phrase that quietly runs through the draft policy is sovereignty, and not in the traditional political sense, but the ability for Kenya to retain oversight of the data, infrastructure, compute resources, knowledge assets, and governance systems that will shape how AI develops within its borders.
For the Kenyan government, AI has become too important for African countries to depend entirely on foreign-controlled infrastructure. Instead, it proposes strengthening domestic capabilities across nearly every part of the AI value chain, from research and language resources to cloud infrastructure, compute capacity, intellectual property, assurance services, and skilled talent.
It also emphasises local value creation, where international partnerships could build domestic expertise rather than deepen dependence on external technology providers.
While public conversations about AI often revolve around tools such as ChatGPT or image generators, Kenya’s draft highlights compute infrastructure, cloud services, interoperable data systems, and connectivity.
The policy proposes expanding affordable and secure compute infrastructure through shared compute resources, sovereign cloud arrangements, stronger connectivity, and closer integration with sustainable energy and water resources. It also stresses the importance of building domestic technical capacity to deploy, maintain, and repair these systems.
The draft cites large-scale investments by the United States, the European Union, the United Kingdom, China, India, and Brazil as evidence that AI leadership is increasingly tied to infrastructure rather than to applications alone.
Its scope also covers other emerging technologies that evolve rapidly, create new governance challenges, or converge with AI. These include technologies such as blockchain, the Internet of Things (IoT), robotics, autonomous systems, extended reality, advanced communications technologies, and quantum-related technologies.
The policy also adopts an unusually broad jurisdiction. It applies not only to organisations operating inside Kenya, but also to foreign companies whose AI systems are used within the country or have direct and foreseeable effects on individuals or public interests in Kenya. Public-sector suppliers, cloud providers, model providers, data annotation companies, AI assurance providers, and technology vendors are all explicitly brought within its scope.
Rather than imposing blanket rules on every AI application, Kenya proposes a risk-based regulatory approach. Higher-risk systems would attract stronger requirements around documentation, testing, monitoring, security, accountability, and human oversight, while lower-risk applications would face proportionate obligations.
To counter the dominance of foreign foundation models, Kenya is launching a dedicated National African Languages and Voice AI Programme. By building governed data commons for Kiswahili and indigenous languages, the state aims to preserve cultural heritage while ensuring digital tools are accessible to non-English-speaking populations.
