WITHIN days of each other this month, legislators in two of Africa’s most influential democracies have placed the governance of new and powerful media technologies squarely on the political agenda. In Nairobi, nominated Senator Karen Nyamu tabled the Artificial Intelligence Bill, 2026 – Kenya’s first comprehensive attempt to regulate the lifecycle of AI systems. In Cape Town, South Africa’s Portfolio Committee on Communications and Digital Technologies announced a landmark roundtable on the regulation of podcasting, scheduled for 24 March 2026. Taken together, the two developments signal a continent-wide inflexion point: African governments are no longer content to watch transformative digital technologies operate in regulatory vacuums.
The convergence is striking in its timing but even more so in its underlying anxiety. Both initiatives are animated by a shared recognition that the digital media landscape has evolved faster than the legal frameworks meant to govern it — and that the consequences of inaction, from deepfake disinformation to unaccountable audio journalism, are becoming increasingly tangible.
SOUTH AFRICA: REGULATING THE PODCAST BOOM
South Africa’s podcasting industry has expanded rapidly on the back of smartphone proliferation, improved broadband penetration, and a creator economy that has empowered independent journalists, commentators, and cultural producers in ways traditional broadcasting never could. Today, South Africans are among Africa’s most prolific podcast consumers, with content spanning investigative reporting, political analysis, lifestyle programming, and Zulu, Xhosa and Sotho-language storytelling.
But the legal architecture governing this ecosystem remains ambiguous. Podcasts distributed over the open internet do not fit neatly within the Broadcasting Act or the ICASA regulatory framework, which were designed for scheduled broadcasting. Questions of accountability — who is responsible when a podcaster spreads defamatory content, incites violence, or operates a commercially significant platform without public interest obligations — have gone unanswered.
It is against this backdrop that Portfolio Committee Chairperson Ms Khusela Sangoni-Diko has convened what she described as “A Multi-Stakeholder Dialogue on Podcasting: Legislators, Policymakers, Regulators, Podcasters, and Industry Stakeholders Charting Balanced Regulation for Sustainable Growth, and Strong Accountability.” The roundtable, which brings together representatives from the Department of Communications and Digital Technologies, independent podcasters, digital platforms, civil society organisations, legal experts, and academic institutions, is explicitly framed not as a precursor to a crackdown but as an exercise in co-regulatory architecture.
“Podcasting represents one of the most exciting developments in South Africa’s digital content ecosystem. It has opened space for diverse voices, languages and perspectives, while creating new opportunities for innovation and economic participation.”
Ms Khusela Sangoni-Diko
Ms Sangoni-Diko’s framing is careful. “Our intention is not to stifle creativity, but to ensure that as the sector grows, it does so within a framework that supports innovation, protects the public interest and expands participation in South Africa’s digital economy,” she said. The committee hopes the engagement will produce consensus points, areas for further work and recommended next steps for policymakers.
What the South African process notably lacks — at least at this stage — is the coercive edge that defines its Kenyan counterpart. The roundtable is consultative. No Bill has been tabled. The emphasis on “co-regulatory models” and “mechanisms for handling complaints” suggests the committee is looking for the lightest effective touch: frameworks that can address public interest concerns without imposing the licensing regimes and content quotas that have long constrained traditional broadcasters.
Press freedom advocates will watch this process closely. South Africa has a constitutionally robust media environment, and any regulatory overreach risks a challenge in court. But the media industry itself has reason to engage constructively. A self-regulatory or co-regulatory outcome — similar to the print media’s Press Council model — could protect podcasters from heavier-handed future intervention while providing a credible complaints mechanism.
KENYA: THE SILICON SAVANNAH’S REGULATORY GAMBLE
If South Africa’s approach is cautious and consultative, Kenya’s is ambitious — some would say alarmingly so. The Artificial Intelligence Bill, 2026, tabled by Senator Karen Nyamu, represents the country’s first comprehensive legislative attempt to govern AI systems from development through deployment to market. Its scope is sweeping, and its enforcement teeth are sharp.
At the bill’s institutional core are three new bodies: an Office of the Artificial Intelligence Commissioner — appointed by the President with parliamentary approval — charged with registering AI systems, conducting audits, investigating compliance breaches and enforcing penalties; an AI Authority tasked with setting national standards and running regulatory sandboxes for innovation testing; and an AI Advisory Council to guide the government on global trends and emerging risks.
The bill establishes a risk-based classification framework modelled closely on the European Union’s AI Act, which entered into force in August 2024. AI systems are categorised by potential harm to health, safety, fundamental rights, the environment, or societal welfare. Those classified as “high-risk” — encompassing AI deployed in healthcare, education, banking, agriculture, law enforcement, employment and public administration — require mandatory registration, conformity audits, and prior approval from the Commissioner before they can be deployed.
“A person shall not develop, deploy or operate a high-risk artificial intelligence system without the approval of the commission.”
Kenya Artificial Intelligence Bill, 2026
The enforcement provisions have attracted the most commentary. Non-compliance with the bill’s provisions carries fines of up to KSh 5 million (approximately $38,000), and certain violations — particularly those involving harmful deepfakes or the unauthorised deployment of high-risk AI — carry prison sentences of up to three years. That criminal dimension sets Kenya apart from most comparable regulatory frameworks globally; the EU AI Act, by contrast, relies on administrative fines rather than incarceration.
Senator Nyamu has cited her own experience as a partial motivation for the bill. The senator — herself a public figure who has faced AI-generated imagery used to harass and defame her — has highlighted deepfakes as a specific harm the legislation must address. The bill mandates clear labelling of all AI-generated content that resembles real persons, places or events, and criminalises the creation or distribution of deepfakes intended to deceive, defame or incite violence. With Kenya’s 2027 General Election approaching, these provisions address a documented and growing threat to democratic discourse.
But the bill has attracted substantial criticism from Kenya’s technology sector. The country’s reputation as Africa’s “Silicon Savannah” — home to a dynamic startup ecosystem spanning fintech, agritech, edtech, and logistics — rests partly on a relatively permissive regulatory environment that has allowed rapid experimentation. Critics argue that the bill’s compliance burden would fall most heavily on small Kenyan developers who build on open-source foundation models and cannot produce the full audit trails of training data and model decisions the bill would require.
A further concern is institutional proliferation. Kenya already has an Office of the Data Protection Commissioner and a Communications Authority with mandates that substantially overlap with what the proposed AI Commissioner would do. Adding three new oversight bodies risks fragmenting regulatory responsibility, creating compliance confusion, and ultimately slowing the very innovation the bill purports to protect.
Perhaps most significantly, the bill’s reach is structurally limited in a way its drafters may not have fully reckoned with. The AI systems that most influence the daily lives of Kenyans — the credit scoring algorithms of global financial platforms, the content recommendation engines of Meta and Google, the large language models underpinning services used in hospitals and schools — were built by companies headquartered in the United States, the United Kingdom, and China. They do not answer to Kenyan courts. A regulatory framework of domestic application, however well-designed, cannot fully govern a global supply chain of AI development.
A CONTINENT FINDING ITS VOICE
Taken together, the South African and Kenyan initiatives reflect a broader African awakening to the governance challenges posed by the digital information economy. For decades, the continent has been a consumer of technologies designed elsewhere, regulated by frameworks developed in Brussels, Washington or London, and distributed through platforms accountable to no African legislature.
That is changing. African governments — sometimes clumsily, sometimes with genuine sophistication — are asserting their authority over digital spaces that shape political opinion, economic behaviour, and social identity. The continent is not moving in one direction: Rwanda has embraced AI adoption aggressively; Nigeria has experimented with social media restrictions; Ethiopia has used internet shutdowns as a tool of political control. Kenya and South Africa, as the continent’s two largest economies by sophistication of their tech sectors, occupy a critical middle ground.
What both the South African podcasting roundtable and the Kenyan AI bill share, despite their different textures, is a recognition that the status quo is untenable. The podcast creator operating outside any accountability framework is not the same as the traditional journalist subject to ethical codes and legal liability. The AI system making loan decisions or shaping electoral information without regulatory oversight is not the same as a human decision-maker operating under law. Regulation is coming. The question is whether it will be designed well.
The risks of poor design are not theoretical. Overbroad content regulation can silence political dissent. Compliance requirements that large corporations can absorb can destroy small African innovators. Regulatory bodies without clear mandates can become patronage vehicles or tools of political interference. Both Kenya and South Africa have seen enough of these pathologies to know the stakes.
What good design looks like, in this space, is contested — but some principles are clear. Regulation should be proportionate to demonstrated harm. It should be transparent in process and accountable in outcome. It should preserve space for African creative and economic entrepreneurship. And it should not treat the digital media and AI sectors as uniquely suspect, to be regulated more harshly than equivalent harms in analogue or established industries.
The South African roundtable, if it genuinely listens, has a chance to model what collaborative, light-touch co-regulation can look like. The Kenyan bill, if it survives the Senate and emerges from scrutiny genuinely improved, could provide the continent’s most ambitious AI governance framework — provided it does not strangle the innovation it is meant to protect.
Africa’s digital future will not be governed from Silicon Valley or Brussels. The continent is insisting on writing its own rules. The quality of what Kenya and South Africa produce in the coming months will say much about whether the continent is ready for the responsibility that sovereignty demands.
