The national government is grappling with how best to regulate generative artificial intelligence, particularly deepfake videos, amid growing concern over disinformation and its influence on the public.
Much of the debate on AI governance has been driven by political urgency and public anxiety, often framing deepfakes solely as a threat to democracy and national security. While these concerns are legitimate, an overly alarmist approach risks overlooking the broader implications of deepfakes for Kenya’s legal system, business environment, and innovation ecosystem.
Recent events illustrate the power and danger of this technology. AI-generated images and videos falsely depicting senior public officials have circulated widely on social media, provoking public outrage, arrests, and heightened political tension.
These incidents demonstrate how easily deepfakes can distort public discourse, undermine trust in institutions, and inflame fragile social dynamics. They also raise a critical policy question: should Kenya regulate deepfakes as an exceptional threat, or govern them as part of a broader, risk-based approach to emerging technologies?
Kenya has no law dedicated to deepfakes. Instead, their regulation is dispersed across multiple legal instruments. The Constitution protects the right to privacy, freedom of expression, and media freedom, while limiting speech that incites violence, hatred, or discrimination. The Data Protection Act governs the processing and misuse of personal data, including images and videos.
The Copyright Act protects creative works from unlawful alteration that prejudices an author’s reputation. Meanwhile, the Penal Code and the Computer Misuse and Cybercrimes Act criminalise the publication of false or misleading information likely to cause panic, reputational harm, or public disorder.
Collectively, these laws provide a foundation for addressing harmful uses of deepfakes. However, they were not drafted with generative AI in mind. They lack a clear definition of what constitutes a deepfake, impose inconsistent thresholds for liability, and are largely reactive rather than preventative.
For businesses, media organisations, and courts, this fragmented framework creates uncertainty around compliance obligations, risk exposure, and enforcement.
The implications for Kenya’s business landscape are substantial. Globally, deepfakes have been used to impersonate company executives, authorize fraudulent financial transactions, and spread false market-sensitive information.
In a rapidly digitising economy like Kenya’s—mobile penetration is high and digital transactions are expanding—such misuse could undermine corporate governance, investor confidence, and consumer trust. SMEs, often lacking sophisticated detection tools, are particularly vulnerable.
Yet focusing exclusively on harm presents an incomplete picture. Deepfakes are not inherently malicious. When responsibly deployed, they can unlock genuine economic and social value. In the creative and entertainment industries, deepfake technology enables advanced visual effects, immersive storytelling, and cost-effective content production.
In education, AI-generated avatars and simulations can personalise learning and make complex or historical concepts more accessible.
In healthcare, realistic virtual patients can enhance medical training without exposing real patients to ethical risk. In legal and forensic practice, similar AI techniques are increasingly used to detect manipulated evidence and protect vulnerable witnesses.
The regulatory challenge, therefore, is one of balance. An overly restrictive or punitive approach risks stifling innovation, discouraging investment, and pushing AI development into informal or unregulated spaces. Conversely, regulatory inaction leaves individuals, businesses, and democratic institutions exposed to serious harm.
Kenya requires a clear, proportionate, and forward-looking framework. This begins with definitional clarity.
A dedicated statute or policy instrument—possibly anchored within the National Artificial Intelligence Strategy—should define deepfakes, distinguish between malicious and legitimate uses, and align sanctions with demonstrable harm and intent. Regulation should focus on misuse, deception, and damage, rather than the technology itself.
Transparency measures could play a central role. Disclosure or labelling requirements for AI-generated content, particularly in political communication and commercial advertising, would promote accountability without banning legitimate creative or commercial applications.
At the same time, businesses should be encouraged to adopt internal AI governance policies addressing risk management, consent, data protection, and reputational safeguards.
Equally important is public and institutional capacity-building. Laws alone cannot address the deepfake challenge if citizens lack the skills to critically assess digital content.
Media literacy, digital verification skills, and public awareness campaigns must complement legal reform. Courts, regulators, and law enforcement agencies also require technical tools and training to authenticate digital evidence in an era where visual proof is no longer self-evident.
Deepfakes reflect a broader reality of Kenya’s digital transformation: technology is advancing faster than law, but not beyond governance.
If approached thoughtfully, deepfakes can support creativity, productivity, and economic growth. If approached through fear or over-regulation, they risk becoming another missed opportunity in Kenya’s innovation journey.
The task before policymakers is not to panic, but to calibrate—protecting citizens and markets from harm while allowing innovation to flourish. In doing so, Kenya can position itself not only as a consumer of emerging technologies, but as a jurisdiction capable of governing them responsibly and competitively.
Brenda Maina is an Advocate of the High Court of Kenya
Unlock a world of exclusive content today!Unlock a world of exclusive content today!
