Kenya’s first dedicated artificial intelligence (AI) bill has stalled. Why? The Senate’s Information, Communication and Technology Committee has asked the Senate not to proceed with the Artificial Intelligence Bill, 2026, saying its penalties were not suited to a sector that is still finding its footing.
According to The Star, the committee was chaired by Trans Nzoia Senator Allan Chesang. The bill was sponsored by nominated Senator Karen Nyamu in early 2026.
The committee also warned that the proposed law, if passed, could drive investors and local developers away and hold back the growth of Kenya’s digital ecosystem, often called the Silicon Savannah.
The Positive Aspects of the Bill
Despite its rejection, the proposal offered many potential gains.
The strongest argument in its favour is legal clarity. Without a dedicated AI law, developers, investors and users must guess which activities are permitted and who bears responsibility when something goes wrong. A defined framework would have removed much of that guesswork, letting businesses plan, invest and build products with confidence about the standards they operate under.
A dedicated law would also have protected citizens in ways existing legislation cannot. The Data Protection Act, 2019 and the cybercrime laws were not designed for AI-specific risks such as algorithmic bias and deepfakes. The Data Protection Act regulates personal data, whereas the bill was designed to protect people from AI systems themselves.
Accountability and oversight are closely linked benefits. Defined liability and transparency duties would have made it easier to hold developers and deployers responsible, while a designated regulator would have stopped individual sectors from improvising their own requirements. A coherent approach would have been easier for companies to follow and for the public to understand.
Trust and investment follow from this. People are more willing to use AI in health, banking and government services when they know guardrails exist, and investors prefer markets where the rules are predictable. Kenyan firms that serve foreign clients would also have found it easier to demonstrate compliance if the national regime aligned with global standards.
Finally, the law could have positioned Kenya as a leader rather than a follower. By acting early, Kenya could have shaped regional norms across Africa instead of adopting rules written elsewhere. It could also have formalised innovation-friendly tools such as regulatory sandboxes, incentives for local startups and support for homegrown AI development, so that oversight and growth advanced together.
These benefits depend on good design. The committee’s concern was that harsh penalties and heavy obligations could burden a young, developing sector, so any fair assessment must acknowledge that a poorly calibrated law can cancel out these advantages.
The Downsides of Rejecting the Bill
The most immediate consequence is a regulatory vacuum. Kenya now has no law written specifically for AI. Questions about liability, bias and misuse remain unanswered, and ordinary Kenyans have limited legal recourse when they are harmed.
This void also exposes citizens to real risks. As AI shapes more decisions about credit, employment and access to services, people affected by biased or opaque systems have no AI-specific route to challenge them. Victims of harmful deepfakes face a similar problem: without targeted offences and enforcement, they must rely on general laws that were drafted before this technology existed.
The absence of legislation also creates uncertainty for businesses and investors. Heavy regulation can discourage investment, but so can the lack of any defined requirements.
Startups, banks and foreign partners do not know what standards will eventually apply, which makes it harder to plan products, secure funding and prove compliance to international clients. Delay does not remove regulatory risk; it only postpones it.
The decision also sets back institutional development. The bill proposed an Office of the AI Commissioner and an Advisory Committee, and, according to Techweez, an AI Authority. It would have given the Commissioner a role advising counties on using AI in healthcare, agriculture and education.
Counties are already using AI in service delivery, often without any governance structure, so rejecting the bill leaves that practice unguided. It also removes the proposed sandboxes, which would have let startups test products under supervision, as Techweez noted.
Which Laws Currently Govern Kenya’s Digital Space?
Kenya’s digital ecosystem is currently governed by the following laws:
- Data Protection Act, 2019: Gives citizens explicit control over their personal information and sets compliance duties for organisations that process, store or transfer personal data across borders. It is overseen by the Office of the Data Protection Commissioner (ODPC).
- Computer Misuse and Cybercrimes Act, 2018 (amended 2025): Targets digital offences including illicit network entry, espionage, breaches and online harassment. Revisions from late 2025 increased penalties and introduced SIM-swap violations, though some free-speech clauses face court challenges.
- Virtual Asset Service Providers Act, 2025: Creates a regulatory framework for digital assets such as cryptocurrencies. It was enacted in November 2025, with proposed procedural rules following in early 2026 to manage financial risks while supporting innovation.
Not in force: The Artificial Intelligence Bill, 2026 would have introduced tiered risk management, compliance auditing, and testing sandboxes under a newly created AI Commissioner. It was halted by a Senate committee.
