Kenya is currently debating the Artificial Intelligence (AI) Bill 2026, sponsored by nominated Senator Karen Nyamu. If passed, it would become the country’s first comprehensive law on how AI systems are developed, used, and regulated.
The goal makes sense, but the structure needs closer review.
The bill proposes three new government bodies to oversee how AI is used in Kenya. The first is an AI Commissioner to act as the main enforcement arm, with powers to investigate complaints, carry out inspections, and penalize those who break the rules.
Second is an AI Authority that would be responsible for setting the standards that developers and companies must meet and would run a controlled testing environment where startups can try out new AI products under supervision before releasing them to the public.
The third body is an Advisory Council that would bring together experts to advise the government on emerging trends and risks as the technology continues to evolve.
Beyond the institutions, the bill also sorts AI systems into categories based on how much harm they could cause, with stricter rules applying to those considered most dangerous.
According to the bill, not all AI will be treated the same. Some uses are banned outright, such as systems that rank citizens by behavior or nudge people into decisions without their knowledge.
Others, like AI used in hospitals, banks, schools, hiring, and policing, are considered high-risk and would need to be registered, regularly reviewed, and kept under human supervision.
Simpler tools, like a basic customer service chatbot, for example, would only need to tell users they are talking to a machine.
On deepfakes and synthetic media, the bill is direct. 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 and up to two years in prison.
Political deepfakes are proposed to carry the same liability.
With Kenya’s 2027 General Elections approaching, this provision addresses a well-documented and growing threat. Individuals affected by automated decisions, whether on loans, jobs, or welfare assessments, would gain the right to a plain-language explanation and a human review.
In February, the High Court issued an order after an urgent petition argued that the absence of AI safeguards threatens fundamental rights. Senator Nyamu has also cited the harassment of female politicians through AI-generated imagery as a driver.
The harder question is whether this law will actually work. Kenya does have a growing AI sector, with local startups building tools for farming, nutrition, education, and business.
The models underpinning most of what Kenyans use daily in banks, hospitals, and government offices were largely built by companies headquartered abroad that do not answer to Kenyan courts or regulators.
No Kenyan official can compel those companies to open their systems for an audit. Where Kenya does have clear jurisdiction is over how AI gets deployed locally, and that is where the real gap is.
When an algorithm rejects a loan application, screens out a job candidate, or flags a medical diagnosis, those decisions are happening in Kenya, affecting Kenyans, with no rules governing them.
That is what this bill is proposing to fix, and on that narrower point, it is on solid ground.
The institutional design is where the bill is most vulnerable to criticism. Kenya already has an Office of the Data Protection Commissioner and a Communications Authority.
Many of the harms this bill targets fall within their existing or extendable mandates. Layering three new bodies on top risks fragmented oversight, regulatory confusion, and a compliance burden that hits local startups hardest.
There is also a practical problem buried in the compliance requirements. Many Kenyan developers do not build AI from scratch. They take existing open-source models, built and trained elsewhere, and adapt them for local use.
Requiring those developers to produce a full audit trail of how the underlying model was trained, what data it used, and how it reached its decisions may simply not be possible. That information is not theirs to provide.
The bill also comes with a price tag. Creating three new government bodies costs money that Kenya’s public budget does not have in abundance, and those costs do not go away after year one.
Businesses operating in high-risk sectors would face registration fees and regular compliance reviews. For large companies, that is manageable. For a small startup, it could be the difference between launching a product and not bothering.
