By 
Kenya is proposing stricter regulation of artificial intelligence systems, with plans to criminalise the deployment of “high-risk” AI tools without prior government approval.
Under the draft Artificial Intelligence Bill 2026, sponsored by Karen Nyamu, individuals or organisations that develop or operate high-risk AI systems without authorisation could face fines of up to KES 5 million (approximately $38,000) or prison sentences of up to three years.
The proposed legislation targets AI applications in sensitive areas such as credit scoring, biometric identification, and healthcare diagnostics—systems that directly influence access to financial services, employment, and essential services.
As AI adoption accelerates across Kenya’s technology ecosystem, these tools are increasingly being used in loan approvals, recruitment processes, fraud detection, and customer service operations. The bill aims to bring such use cases under closer regulatory oversight.
A key issue raised by stakeholders is how “high-risk” AI will be defined in practice. A broad classification could extend regulatory requirements to sectors including finance, healthcare, education, and even general-purpose AI models integrated into everyday applications.
The proposed framework would also introduce personal liability for company executives, meaning directors could be held accountable for approving the deployment of non-compliant AI systems.
For startups, which often rely on rapid development cycles and third-party APIs from global providers, the requirement for prior approval could create delays in product launches and increase compliance costs.
The bill further proposes the establishment of an AI commissioner with authority to classify systems, grant approvals, and maintain a public registry of AI tools in operation. Regulators would also be empowered to inspect AI systems, datasets, and related documentation, expanding oversight into how digital products are developed and deployed.
This approach marks a shift toward stricter enforcement, extending beyond administrative penalties into criminal sanctions. It contrasts with regulatory models in regions such as the European Union and the United Kingdom, where AI governance frameworks typically rely on audits, compliance requirements, and financial penalties rather than imprisonment.
The proposed inspection powers could also raise concerns for companies handling proprietary data or developing in-house AI models, as they may need to balance regulatory compliance with the protection of intellectual property and trade secrets.
Commenting on the development, Mike Olukoye, a technology legal expert based in Nairobi, noted that stronger oversight is necessary as AI systems increasingly influence critical decisions affecting individuals’ access to opportunities and services.
He added that early regulatory intervention could help mitigate risks and prevent large-scale harm as AI adoption continues to grow.
The proposed legislation reflects Kenya’s effort to establish a structured regulatory framework for artificial intelligence, balancing innovation with the need to protect consumers and maintain trust in emerging technologies.
Comments