Kenya's first national AI policy keeps citizens' data under Kenyan law

Kenya's first national AI policy flips the default on AI data: anything generated inside the country must be governed under Kenyan law, on infrastructure the government plans to build and own domestically. The Kenya Artificial Intelligence and Other Emerging Technologies Policy, 2026 was released for public participation on July 21, 2026, per a notice carried by local outlets and reporting by The Star. The consultation window closes August 4.

The policy requires AI platform providers to ensure that data generated in Kenya is governed under Kenyan laws, and binds any organization processing personal information to the existing Data Protection Act. In plain terms, the data an AI system produces or trains on inside Kenya now has a default Kenyan jurisdiction for the first time. The same section pairs that jurisdictional flip with a public-investment commitment: state funding for local data centers, cloud infrastructure, and high-performance computing, framed in the policy text as "sovereign digital infrastructure". In practice, the phrase means compute and storage built and run inside Kenya rather than rented from foreign-owned cloud regions.

The policy creates two new bodies: a National AI and Other Emerging Technologies Council to coordinate across ministries, regulators, academia, and the private sector, and a Kenya AI Safety Institute to audit high-risk AI systems for bias, security, and ethical compliance. The framework also commits to AI tools developed in Kiswahili and other Kenyan languages, and sets a short list of safeguards: transparency about when a person is interacting with AI, the right to question or seek review of algorithmic decisions, and protections for children, persons with disabilities, and other vulnerable groups.

The policy names five priority sectors for AI adoption: agriculture, healthcare, education, financial services, and public administration, with concrete use cases ranging from drought, pest, and crop-disease prediction to faster diagnosis of tuberculosis and cervical cancer, personalized learning, fraud detection in finance, and service-delivery automation in government. That list is a standard AI-rollout menu, useful as scope but not the mechanism. The mechanism is the default jurisdictional flip underneath it.

The policy does not arrive alone. It runs in parallel with the Artificial Intelligence Bill, 2026, a Senate proposal that is not yet law, and with the Kenya National AI Strategy 2025-2030, the government's broader five-year roadmap for AI capacity. The three tracks have different legal force: the strategy is a roadmap, the policy is a ministerial framework that takes effect after the consultation closes, and the bill is a Senate proposal. Treating them as one regulatory event would obscure which lever actually changes which behavior.

Written comments on the policy go to aipolicy@moict.go.ke by August 4, 2026, with a copy to legal@moict.go.ke, on a template issued by the committee chaired and hosted at the Principal Secretary's office, Telposta Towers in Nairobi. The "sovereign digital infrastructure" language is the policy's own framing, and the public comment period is the point at which that phrase gets pinned down: which data counts, which providers are in scope, what "domestic" means for cloud and compute, and how the new AI Safety Institute relates to existing regulators.

The backdrop is regional positioning. ICT Cabinet Secretary William Kabogo attended the 2026 World AI Conference in Shanghai on July 17, 2026, four days before the public-participation notice. The two events are not the same story, but they bracket the same claim: Kenya intends to be a rule-maker for AI on the continent, not a rule-taker. Whether the commitment to local data centers, cloud, and high-performance computing turns into line-item budget or stays at policy-paper level is the next concrete thing to watch.