PressBridge

On any given day in Nairobi, you can walk into a small office where a developer is training a model to spot fraud, a clinic is testing a tool that helps triage patients faster, or an insurer is trying to automate claims so customers stop waiting for weeks. The energy is real, and the ambition is loud. Africa is not short on talent, ideas, or urgency. 
What keeps many of these breakthroughs stuck in pilot mode is not a lack of innovation, but the absence of clear, trusted and workable policy that turns promising experiments into systems people can rely on.
Across the continent, the conversation on artificial intelligence has moved from curiosity to necessity. Governments are drafting national strategies, and private sector players are investing in use cases that touch real lives, from healthcare to finance. Development partners are positioning AI as a driver of productivity and growth, yet adoption remains uneven, fragmented in execution, and often trapped in small trials that never reach national scale. The missing link is policy.
Policy is how a society decides what is safe, what is fair, what is allowed, and what is worth investing in. It is how citizens learn to trust systems that increasingly shape decisions about loans, jobs, medical care, and public services. It is also how investors decide whether to back a product today, or wait until the rules are clearer. 
In the world of business, AI does not scale in isolation but inside the guardrails, incentives, and direction set by policy. In a continent where markets are diverse, regulations are evolving, and infrastructure gaps still exist, policy is not just an enabler but the main accelerator.
There are three ways policy determines whether AI becomes a real tool for development. The first is certainty that enables businesses to invest with clarity. When rules around data governance, privacy, accountability, and ethical use are defined and predictable, innovators can build with confidence and investors can commit long term capital without fearing sudden regulatory shocks. 
The second is alignment that guides AI toward problems that keep citizens up at night, whether that is access to healthcare, food security, inclusion in financial services, or more responsive government. Policy can direct innovation toward national priorities and stop it from becoming scattered experimentation that looks impressive but solves little. 
The third is trust, especially in data managment. AI depends on data, often sensitive data. If people feel exposed, exploited, or unheard, adoption will stall. Strong frameworks for privacy, security, redress, and oversight help citizens believe that technology can serve them rather than watch them.
Kenya is a useful case study because it shows what it looks like when a country tries to move from talk to law. In March 2025, the Ministry of ICT published a National AI Strategy for 2025 to 2030, with an ambitious investment plan and a clear recognition that Kenya needs comprehensive regulation to address ethical implications and potential harms. That strategy matters because it signals intent. 
Yet there remain to be settled the many questions investors and citizens ask every day. Who is responsible when AI causes harm, how is data collected and stored, what standards apply to systems used in hospitals, banks, or public services. What protections exist for children, patients, and vulnerable communities. These questions are important because they decide whether AI is trusted and used.
That is why the Artificial Intelligence Bill 2026, currently undergoing public participation, is such an important moment. The Bill proposes a risk classification approach that echoes global thinking, grouping AI systems from unacceptable risk to minimal risk. This kind of structure can be helpful because it avoids treating every AI tool as equally dangerous. A chatbot that answers customer queries is not the same as a system that influences medical decisions or credit scoring. A good framework recognises this difference and allocates stronger oversight to high stakes use cases.
Still, the African context requires care. If rules are copied without adaptation, they can punish local innovators while leaving the most powerful actors comfortable. A developer in Nairobi is not asking to escape accountability but for a framework that is designed with her reality in mind. She is asking for regulation that protects citizens without raising compliance costs so high that only large foreign firms can compete. If policy becomes a gate that only the rich can open, we get regulation without justice and innovation without local ownership.
This is where business leadership becomes central. The private sector cannot sit back as a passive recipient of policy and then complain about outcomes. In the AI era, business leaders must be co- creators because they stand at the frontlines of implementation. They know where systems fail, where inefficiencies live, and where AI can deliver measurable value fast. They can provide evidence on what works and what breaks, help regulators understand the practical impact of proposed rules on small firms and on adoption in sectors like health, insurance, agriculture, and public administration. 
If Africa wants AI that improves lives, three shifts are urgent. We must move from policy announcements to policy execution with clear implementation plans, timelines, and accountability mechanisms. We must move from siloed efforts to ecosystem collaboration, where government, civil society, academia, and business build a shared understanding of risk, opportunity, and enforcement. And we must move from short term pilots to scalable systems.  
When policy is right, AI can reduce the cost of healthcare and expand access. It can help farmers respond to climate shocks and improve yields. It can make financial services cheaper and more inclusive. It can cut waste in government and speed up services that currently exhaust citizens. It can create jobs in data work, software development, implementation, auditing, and compliance. It can build African champions, not just African consumers.
But when policy is wrong, adoption becomes fragmented, trust erodes, and the value remains unrealised. The continent then becomes a market for tools built elsewhere, governed by rules written elsewhere, and optimised for priorities that may not match our own.

The author is the Group Managing Director, Smart Applications International,  a leading health-tech ICT solutions provider.

By Harrison Muiru