The proposed Kenya Artificial Intelligence and Other Emerging Technologies Policy says individuals “are entitled to know when an AI system is acting on them” and requires organisations to disclose when customers are communicating directly with AI. 

If adopted, the requirements could affect banks using AI in credit decisions, insurers processing claims, telecoms operating customer service chatbots, employers screening job applicants, hospitals using clinical decision support tools, and media companies publishing AI-generated content. 

“Where a person interacts directly with an AI system in place of a human, where an automated system makes or materially shapes a decision that significantly affects their rights, access to services, or legal interests, or where content presented to the public has been generated or substantially altered by AI, that fact shall be disclosed in a clear and accessible manner,” the policy said.

The measures mirror a growing global push for transparency as AI systems become embedded in everyday services. The European Union’s AI Act requires users to be informed when interacting with certain AI systems, while providers of AI-generated or manipulated content must clearly disclose it in many cases. 

In the US, states including California have passed laws requiring political advertisements containing AI-generated content to carry disclosures, and several federal agencies have warned companies against misleading consumers about their use of AI. 

The disclosure requirement is one of the clearest consumer-facing provisions in the 226-page draft policy, which seeks to move Kenya beyond promoting AI adoption to regulating its use across the economy. The government says the framework is intended to balance innovation with safeguards against discrimination, opaque decision-making, and abuse. 

Rise of generative AI

The proposal comes at a time when Kenyan businesses, such as banks and insurers, are embedding generative AI into customer support, fraud detection, marketing, coding, and internal operations. Most of these are done without the customers’ knowledge.

The policy proposes that AI deployed should be “explainable, auditable, and subject to human oversight” throughout its lifecycle. It also says that humans should retain control over AI systems that affect individual rights, safety, or access to essential services.

“Humans shall retain meaningful control over systems that affect individual rights, safety, or access to essential services,” the draft stated, adding that high-risk systems should include mechanisms allowing human review, intervention, and clear allocation of responsibility.

The transparency push forms part of Kenya’s effort to become a regional centre for AI governance. In May, the government backed the Africa Forward Declaration, which calls for greater investment in African compute infrastructure, sovereign data systems and locally developed AI models, while also securing the right to host the 2027 Responsible AI in the Military Domain (REAIM) summit, the first time the global forum will be held in Africa. 

The policy proposes creating a national AI council and an AI office to coordinate regulation across government while promoting domestic research, infrastructure, and investment. 

While the draft does not specify whether every chatbot interaction would require a notification, how companies should determine when AI has “materially shaped” a decision, or what penalties organisations could face for failing to disclose AI use, if adopted, those details are expected to be addressed through implementing regulations.

True scale demands moving beyond surface-level integrations to robust execution. We’ve filtered the noise out of Moonshot 2026, optimising the conference strictly for high-calibre connections between startup founders, global financial operators, enterprise leaders and individuals rewiring Africa’s technical frameworks.
Get 20% off Early Bird tickets for a limited time.