Kenya's AI Bill Sets a Ksh 5 Million Fine for Deepfakes and Forces Businesses to Audit AI Job Displacement

The Artificial Intelligence Bill 2026 is before the Senate. It creates an AI Commissioner, bans harmful deepfakes with real penalties, and requires workforce impact assessments before deploying AI that could cost people their jobs.

Kenya’s Senate is debating a law that will make distributing AI-generated deepfakes a criminal offence carrying up to two years in prison or a Ksh 5 million fine - and it will require any business deploying AI that could displace workers to prove they assessed the risk first.

The Artificial Intelligence Bill 2026, sponsored by nominated Senator Karen Nyamu, was reported on by Daily Nation and The Standard. It is the most comprehensive attempt yet to regulate AI in Kenya, and it will affect every business using AI tools - not just tech companies.

What the Bill Actually Says

On deepfakes: The Bill proposes fines of up to Ksh 5 million or prison terms of up to two years for distributing harmful AI-generated content using another person’s likeness without consent. Political deepfakes carry the same liability. The provision is designed to address AI-generated impersonation fraud and AI-powered election interference ahead of Kenya’s 2027 General Election.

On job displacement: The Bill mandates that any business deploying AI systems likely to affect employment must carry out a workforce impact assessment before deployment. The assessment must identify which roles are at risk and include a mitigation plan - retraining, redeployment, or phased transition - for affected workers. Developers of those AI systems share responsibility for building in safeguards.

On governance: Central to the Bill is a new Office of the Artificial Intelligence Commissioner, a standalone regulatory body that will oversee AI innovation, investigate complaints, enforce penalties, and issue sector-specific AI guidelines. This is separate from the Office of the Data Protection Commissioner, though the two will work in parallel.

On high-risk systems: The Bill introduces a risk-based classification. High-risk AI systems - those used in credit scoring, hiring, healthcare, law enforcement, or education - face stricter governance, transparency, and record-keeping requirements. Businesses running these systems will need to be ready to explain how decisions are made.

What This Means for a Kenyan Business Owner Right Now

The Bill has not passed yet. But the direction is clear, and businesses that build good practice now will face far less disruption when it does.

If you run AI customer service or chatbots: Ensure your system has a clear human escalation path. The “human-in-the-loop” requirement in the Bill means fully automated decisions on matters affecting customers’ rights or finances will face scrutiny. Build that escalation in now.

If you use AI for hiring or performance assessment: Workforce impact assessments are coming. Start documenting the rationale behind any AI-assisted HR decision today. An audit trail is the minimum defence if a complaint is made.

If you create content using AI tools: Understand that using AI-generated likenesses without consent is about to carry real criminal penalties. This applies to marketing material, social media content, and internal communications. Your content team needs a clear policy.

If you are considering automating roles: The Bill does not ban automation. It requires you to plan for it responsibly - which is good practice regardless. Businesses that automate with a clear retraining pathway will attract better talent and face fewer internal disruptions than those that automate without warning.

A logistics company in Mombasa we spoke with recently automated its dispatch scheduling, cutting the role from three coordinators to one. They did it with six months’ notice, a reskilling programme in route data analysis, and retained all three staff in expanded roles. That is the model the AI Bill is trying to encourage.

The Most Practical Thing to Do Before This Bill Passes

Map your AI use now. List every AI tool your business currently uses - this includes ChatGPT, WhatsApp automation, AI-assisted accounting, customer chatbots, and any AI-powered screening tools. For each one, note: what decision does it inform? Who does it affect? Is there a human review step?

That map is the foundation of your compliance position when the AI Commissioner’s office opens its doors.

AI Consultancy Kenya can run that mapping exercise with you and help you identify which systems will fall into the high-risk category under the Bill’s current draft. WhatsApp us at 0711 344 702 to get started.