Kenya’s Senate is debating the Artificial Intelligence Bill, 2026, a proposed law aimed at regulating artificial intelligence, criminalizing the misuse of AI-generated content, and establishing a national AI regulator. The bill comes as concerns grow over deepfakes and digitally manipulated content ahead of future elections.
The legislation, sponsored by nominated Senator Karen Nyamu, would introduce comprehensive rules for the development and deployment of AI technologies across the country. Its introduction follows a February 6 High Court order responding to a petition arguing that the lack of safeguards for high-risk AI systems threatens fundamental rights, including privacy and equality.
A central feature of the bill is the creation of the Office of the Artificial Intelligence Commissioner, which would serve as the national authority for AI regulation. The commissioner would have powers to investigate complaints, impose penalties, and require companies to modify algorithmic systems to ensure compliance with the law.
The bill establishes criminal liability for individuals or organizations that generate or distribute AI content using someone else’s image, voice, or likeness without consent, where such content causes misinformation, harm, defamation, or privacy violations. Convictions could attract fines of up to 5 million Kenyan shillings, a two-year prison sentence, or both.
Technology providers creating tools capable of manipulating voices, images, or likenesses would be required to obtain clear consent from affected individuals, and all AI-generated content must be clearly labeled as such. The law extends to political campaigns, meaning fabricated videos portraying opponents could result in criminal charges.
Provisions in the bill also protect citizens affected by automated decision-making. Individuals would be entitled to human review of AI-based decisions, the opportunity to express their views, and the right to challenge outcomes, including in employment screening, loan approvals, welfare distribution, and insurance assessments.
Companies deploying AI systems would be required to explain how their technology operates, its purpose, limitations, and the degree of automation involved. Firms must also disclose measures taken to mitigate bias, with penalties of up to 1 million shillings for non-compliance.
The bill proposes a four-tier classification system for AI technologies based on risk. High-risk systems, including those used in healthcare, education, agriculture, finance, security, employment, and public administration, would face stricter oversight. The commissioner would maintain a public register of all high-risk AI systems, including those used by county governments.
Additional provisions include the establishment of an Advisory Committee comprising representatives from the Data Protection Commissioner, the National Commission for Science, Technology and Innovation, and the Council of Governors. Regulatory sandboxes would also be created to allow innovators to test new AI technologies under supervised conditions.
