Kenya just did something most African governments have only talked about. While the continent's tech policy conversations have long been dominated by aspirational frameworks and donor-funded workshops, Kenya's Senate last week introduced the AI Bill 2026, a concrete, enforceable piece of legislation that would make it the first country in East Africa with a comprehensive law governing how artificial intelligence is developed, deployed, and regulated. Sitting here in Harare, watching this unfold, the reaction should not be admiration from a distance. It should be urgent.

The bill, sponsored by nominated Senator Karen Nyamu, proposes three new government institutions to oversee AI across the country. The first is an AI Commissioner, a dedicated enforcement authority with the power to investigate complaints, conduct inspections, and penalise violations. The second is an AI Authority responsible for setting development standards and running a supervised regulatory sandbox where startups can test new AI products before public release. The third is an Advisory Council of technical experts tasked with keeping the government ahead of emerging risks.

Beyond the institutional architecture, the bill introduces a tiered risk classification system. Some AI applications would be banned outright; systems that rank citizens by behaviour or manipulate decision-making without consent are explicitly prohibited. High-risk deployments in hospitals, banks, schools, hiring processes, and policing would require formal registration, regular audits, and continuous human oversight. Lower-risk tools, like a basic customer service chatbot, would simply need to inform users they are interacting with a machine.

On deepfakes and synthetic media, the bill is unambiguous. Anyone who generates or distributes AI-created content using another person's image, voice, or likeness without consent faces fines of up to KES 5 million (approximately USD $38,500) and up to two years in prison. Political deepfakes carry the same penalty, a provision that matters enormously with Kenya's 2027 general elections approaching, and one that Senator Nyamu has directly linked to the harassment of female politicians through AI-generated imagery.

Citizens would also gain the right to a plain-language explanation and a human review whenever an automated system makes a consequential decision about them, a loan, a job application, or a welfare assessment.

The bill is not without legitimate criticism. Kenya already has an Office of the Data Protection Commissioner and a Communications Authority. Adding three new bodies on top of them risks creating fragmented oversight, regulatory confusion, and compliance costs that hit local startups the hardest, the very innovators this framework is meant to protect.

There is also an enforcement ceiling that no piece of legislation can wish away. Most of the AI that Kenyans encounter daily, in banking apps, hospital systems, and government platforms, was built by companies headquartered in San Francisco, London, or Beijing. No Kenyan official can compel those companies to open their systems for a local audit. What Kenya can govern is how AI is deployed on its soil and how it affects its citizens. On that narrower but essential point, the bill is on solid ground.

The compliance requirements also raise a practical problem for local developers. Many Kenyan engineers do not build AI from scratch, they take existing open-source models, trained elsewhere, and adapt them for local use. Requiring those developers to produce a full audit trail of how an underlying model was trained may simply not be possible. That information is not theirs to provide.

Here is where the conversation becomes personal. Zimbabwe is not watching Kenya from a standing start. President Emmerson Mnangagwa launched Zimbabwe's National AI Strategy 2026–2030 at the Parliament Building in Harare, CurrencyLive, setting an ambition to transform the country from a resource-dependent economy into a knowledge-driven one. The strategy rests on six pillars: AI talent and capacity development; national AI infrastructure and computational sovereignty; AI adoption and service transformation; governance, ethics, and regulation; research, development, and innovation; and international collaboration and diplomacy.

Critically, the strategy introduces a national AI regulatory sandbox called the Innovation Crucible, which will allow startups to test AI products under temporary regulatory flexibility, with the first cohort expected to include five to seven fintech and telecoms companies. There is also a national AI literacy campaign branded Nzwisiso.ai, which aims to reach 60% of Zimbabwe's adult population by 2030, one of the more ambitious public awareness targets in any African AI strategy to date.

What Zimbabwe does not yet have is what Kenya is now building: binding, enforceable law. A strategy is a vision. A bill is a commitment. The gap between the two is where citizens get hurt: where an algorithm rejects a loan with no explanation, where a hiring tool screens out candidates based on criteria nobody can examine, and where a deepfake circulates unchallenged because there is no law to invoke.

What makes Kenya's approach particularly instructive is its governance philosophy. Rather than borrowing frameworks from the European Union or the United States, Zimbabwe's own strategy's governance pillar is anchored in Ubuntu, the Bantu philosophical tradition whose central tenet "I am because we are," places collective well-being above individual or commercial interest. Kenya's bill, meanwhile, focuses on individual rights to explanation and human review. There is a productive tension between these two approaches that African policymakers should be interrogating openly rather than importing frameworks wholesale from elsewhere.

Zimbabwe joins a growing number of African nations, including Ghana, Kenya, and South Africa, that are formalising national AI strategies to guide the ethical deployment of intelligent technologies. But strategy documents and enforceable legislation are different animals entirely. Kenya is moving from aspiration to law. The rest of the continent, Zimbabwe included, needs to close that gap, and fast. AI is not waiting for our parliaments to catch up.

The AI Bill 2026 is imperfect. It may be revised significantly before it passes, or it may stall altogether. But the act of introducing it, of forcing the legislature to define what is acceptable and what is not, of putting deepfake penalties on paper and giving citizens the right to challenge automated decisions, that is governance in action. That is a government taking its citizens seriously.