The Artificial Intelligence Bill, 2026: What Kenya’s Proposed AI Law Would Actually Require

A bill that would create Kenya’s first dedicated artificial intelligence regulator is currently before the Senate. The Artificial Intelligence Bill, 2026 has not been enacted, and its provisions could still change before it does, but it is far enough along, and specific enough about what it would require, that technology companies operating in Kenya should understand what it currently proposes rather than wait for a final text to start thinking about compliance.

Where the Bill actually stands

The Artificial Intelligence Bill, 2026 (Senate Bills No. 4 of 2026) was sponsored by Senator Karen Nyamu, published on 19 February 2026, and had its First Reading in the Senate on 2 April 2026. As of the most recent independent legislative tracking, it has passed to Second Reading, with Committee of the Whole House, Third Reading, and presidential assent still to come before it could become law. It has not been enacted, gazetted, or withdrawn. Anything below describes what the Bill as currently drafted proposes, not settled Kenyan law, and any of it could be amended before the Bill passes, if it passes at all.

An Office of the Artificial Intelligence Commissioner

The Bill’s central institutional proposal is a new State Office, the Office of the Artificial Intelligence Commissioner, established under Article 260(q) of the Constitution. The Commissioner would be recruited through the Public Service Commission, nominated by the President, and approved by Parliament, serving a five-year renewable term with defined grounds for removal. Proposed functions include enforcement of the Act, conducting risk assessments and conformity audits of AI systems, maintaining a register of high-risk AI systems, operating regulatory sandboxes for AI innovation, and handling public complaints. This is a materially different regulatory model from Kenya’s existing tech-adjacent regulators: rather than folding AI oversight into the Communications Authority or the Office of the Data Protection Commissioner, the Bill proposes a standalone body built specifically around AI risk.

A risk-based structure, not a blanket rule

Consistent with the general international direction of AI regulation, the Bill’s stated purpose is to regulate and govern AI in Kenya “ensuring ethical, transparent, and accountable use while fostering innovation and safeguarding human rights, data protection, and public welfare.” The register of high-risk systems and the conformity-audit function both point toward a tiered approach, where the compliance burden scales with the risk the system poses, rather than a single set of rules applied uniformly to every AI application. For startups building lower-risk tools, this could mean a comparatively light touch; for systems used in credit scoring, employment decisions, or public services, the audit and registration obligations are likely to be substantive if the Bill passes in anything close to its current form.

A specific deepfake and synthetic media offence

One of the Bill’s more concrete, criminally-enforceable provisions targets non-consensual use of a person’s image, voice, or likeness through AI-generated synthetic media, where doing so causes harm, spreads misinformation, or is defamatory. As drafted, this carries penalties of up to KES 5 million or two years’ imprisonment. This is worth flagging separately from the Bill’s broader institutional architecture because, unlike the Commissioner’s office and the risk-tiering framework, it is the kind of provision that could plausibly survive largely intact even if other parts of the Bill are substantially rewritten during Senate debate, since deepfake harms are a comparatively uncontroversial target of regulation internationally.

How this relates to Kenya’s other AI documents

The Bill is one of three related but distinct Kenyan AI instruments currently in play, and conflating them is a common mistake. The Kenya National Artificial Intelligence Strategy 2025-2030, launched in March 2025, is a non-binding strategy document, not a bill and not enforceable. Separately, the Ministry of Information, Communications and the Digital Economy has published a Draft Kenya Artificial Intelligence and Other Emerging Technologies Policy for public consultation, which we cover in a companion piece on Kenya’s Draft AI and Emerging Technologies Policy. That policy document and this Bill are being developed in parallel by different arms of government, and civil-society submissions on both processes have specifically flagged the need for the two to be coordinated rather than developed in isolation. A business tracking Kenya’s AI regulatory direction needs to watch both, not just one.