Kenyan Senate introduces first AI regulation bill

Kenya Senate has introduced the country first comprehensive artificial intelligence regulation bill, marking a significant step toward governing emerging technologies while sparking debate among technology experts, legal analysts and civil society groups.

The Artificial Intelligence Bill 2026, sponsored by Nairobi Senator Karen Nyamu, proposes a wide-ranging framework to oversee the development and use of AI systems. The legislation aims to position Kenya as a leader in digital governance within Africa, but critics warn that some provisions could slow innovation in the country rapidly expanding technology sector.

Proposed regulatory structure

The bill establishes a three-tier governance system consisting of the Office of the Artificial Intelligence Commissioner, an Artificial Intelligence Authority and an Artificial Intelligence Advisory Council. These institutions would oversee compliance, develop national AI strategies and advise on emerging risks.

The Office of the Artificial Intelligence Commissioner would serve as the main enforcement body, with powers to inspect systems, conduct audits and investigate violations. It would also have authority to ban AI technologies considered high-risk, including systems designed to manipulate human behaviour or produce harmful deepfakes.

Supporters say the framework is necessary to ensure accountability and public trust as AI adoption grows across sectors such as healthcare, finance and education.

Risk classification and compliance rules

The legislation introduces a risk-based classification model for AI systems. Applications deemed to pose unacceptable risk would be prohibited, while high-risk systems would face strict oversight, including mandatory registration, annual impact assessments and human supervision.

High-risk categories include AI used in law enforcement, banking, education and critical infrastructure. Systems considered low or minimal risk, such as basic chatbots, would be subject to lighter transparency requirements.

The bill also requires developers and organisations to disclose when content is generated by artificial intelligence, particularly when it resembles real individuals or events.

In addition, the proposed law grants citizens new rights, including the right to explanation for decisions made by AI systems and the right to request human review in cases where automated decisions significantly affect them.

Penalties raise concerns

The bill outlines strict penalties for non-compliance, including fines of up to KSh 5 million or imprisonment for up to two years for serious violations such as misuse of AI or failure to meet regulatory requirements.

Legal experts warn that these penalties could create a chilling effect, particularly for startups and independent developers. Smaller firms may struggle to navigate complex compliance processes or absorb potential legal risks, potentially limiting participation in the sector.

Critics argue that such measures may favour large multinational companies with greater financial and legal resources, while discouraging local innovation.

“The intention is clear, but the penalties may be too harsh for a developing ecosystem,” said a Nairobi-based technology analyst.

Overlap with existing institutions

Industry stakeholders have also raised concerns about duplication of roles with existing regulatory bodies such as the Office of the Data Protection Commissioner and the Communications Authority of Kenya.

They argue that introducing additional institutions could create a fragmented regulatory environment, increasing costs and administrative burdens for businesses seeking to deploy AI solutions.

Some experts say a more coordinated approach using existing frameworks could achieve similar goals without adding complexity.

A significant portion of the bill targets the misuse of synthetic media, including deepfakes. It proposes criminal penalties for creating or distributing AI-generated content intended to deceive, defame or cause harm.

However, critics say the definitions of harmful content remain unclear, raising concerns about potential misuse of the law to restrict legitimate expression such as satire or political commentary.

Free speech advocates have called for clearer distinctions between malicious content and creative or educational uses of artificial intelligence.

Impact on startups and infrastructure

The bill emphasis on data sovereignty and local infrastructure has also sparked debate. While aimed at strengthening national control over data, experts say the requirement could increase operational costs for startups that rely on global cloud services.

Kenya technology ecosystem depends heavily on international platforms and open-source tools. Developers warn that strict localisation and audit requirements may be difficult to meet, potentially limiting access to advanced AI technologies.

To address innovation concerns, the bill proposes regulatory sandboxes that would allow startups to test AI products in controlled environments with reduced compliance requirements during development stages.

As Senate begins deliberations, stakeholders are calling for amendments to ensure the legislation supports both innovation and public protection.

The outcome of the debate could shape the future of artificial intelligence development in Kenya and influence regulatory approaches across the region.