Across Africa, AI regulation is moving from principles into draft legislation. Kenya’s 2026 Artificial Intelligence Bill is an example of that shift: a proposal to create a formal oversight structure and define obligations for higher‑risk AI use.

Primary source: the Bill text

The Bill is published on Kenya Law’s legislative platform, which provides official access to draft instruments: Kenya Law: The Artificial Intelligence Bill, 2026. Media summaries have also discussed the proposed watchdog model (example: ITWeb Africa).

Why this is a big deal

A dedicated oversight body changes the adoption calculus. It creates a single point of interpretation for risk, compliance, and enforcement — which can reduce ambiguity for businesses, but also raise the compliance bar.

For high‑impact sectors such as finance, healthcare, and public services, a watchdog model can accelerate three expectations:

• Registration: visibility of what AI systems exist and where they are used.
• Human oversight: clear accountability for decisions with real‑world consequences.
• Transparency: disclosure that AI is in use, plus documentation of intended purpose and limits.

What organisations should do now

If you operate in Kenya (or plan to), start with operational readiness:

• Model inventory: map all AI uses, including “hidden AI” in vendor tools.
• Data mapping: identify sensitive data flows and retention.
• Risk assessment templates: standardise how you document impact and mitigation.
• Human decision pathways: define who can override, audit, and remediate.

Legislation may change as it progresses, but the direction is clear: more formal oversight, higher documentation expectations, and more explicit accountability.