🚨 Senate committee has rejected the Artificial Intelligence Bill, 2026, warning that its penalties could harm the sector more than they protect the public. This is the Artificial Intelligence Bill, 2026, sponsored by nominated Senator Karen Nyamu. Here is what it means. The Bill set out to regulate how artificial intelligence is built and used in Kenya. Its main features were: An AI regulator. It proposed an Office of the AI Commissioner to register AI systems, audit them and investigate breaches, supported by an AI Authority to set standards and run testing sandboxes. A risk-based approach. Systems would be sorted by risk. Some uses would be prohibited, high-risk uses in areas like healthcare, education and law enforcement would require registration, annual impact assessments and human oversight, and low-risk tools like chatbots would need only to disclose that a user is dealing with a machine. Penalties. Non-compliance could attract fines of up to KSh 5 million, or a share of annual turnover, and certain offences, including harmful deepfakes, could carry up to two years in prison. The Senate committee on Information, Communication and Technology rejected the Bill. Its concern was that these provisions are too punitive, and that heavy fines, criminal liability and strict compliance duties could discourage innovation, scare away investors, and push local developers out of the market. This leaves a gap. Kenya still has no dedicated AI law. For now, AI is governed mainly by the Data Protection Act, 2019, which covers personal data and automated decision-making, but not the wider questions of AI safety, accountability and deepfakes that the Bill tried to address. The legal dispute worth noting. On one side, regulating a young industry too heavily can raise costs and drive developers and investment elsewhere, which is the committee's concern. On the other, leaving AI largely unregulated leaves real risks to privacy under Article 31, to fair treatment in automated decisions under Article 47, and to protection from discrimination under Article 27, with limited recourse. Both the over-regulation and the gap carry a cost. Key takeaway: rejecting the Bill is not a decision that AI should go unregulated. It is a decision that this design was not the right one. How to regulate AI remains open. Should Kenya regulate AI now and refine as it learns, or wait until the technology and its risks are clearer? 👉 Follow @Lexken_EMSLaw for practical legal insights on technology law and artificial intelligence regulation.