"These so-called 'smart city' surveillance products are anything but smart for those at risk of being tracked and targeted by them. The recording and retaining of facial images of individuals in public spaces without their consent is not legal, necessary or proportionate — it is the latest tool used by governments to invade citizens' privacy and stifle freedom of expression."— Wairagala Wakabi, Executive Director, CIPESA & Co-Author, IDS Smart City Surveillance in Africa Report, March 2026
Kenya's Senate is actively deliberating the Artificial Intelligence Bill 2026, sponsored by nominated Senator Karen Nyamu following her own experience of AI-generated deepfake imagery circulating online — the bill proposes the creation of an independent Office of the Artificial Intelligence Commissioner, an AI Authority, and an AI Advisory Council, with the Commissioner empowered to inspect AI systems, impose penalties, investigate complaints, and manage regulatory sandboxes. The bill introduces criminal liability — fines of up to KES 5 million (approximately $40,000) and prison sentences of up to two years — for anyone generating or distributing AI content using another person's image, voice, or likeness without consent where such content causes harm, defamation, or misinformation; it also grants Kenyans the right to a plain-language explanation of automated decisions affecting them, and the right to human review of such decisions. The Senate simultaneously issued a directive to the Ministry of Information, Communication and Digital Economy to draft a national AI policy covering research, ethical guidelines, regulatory sandboxes, public-private skills partnerships, and AI integration into Kenya's education curriculum — giving the bill both legislative and executive momentum at a moment when the High Court of Kenya had already issued a February 2026 order demanding AI regulation following an urgent petition.
A TechCabal investigation published on 19 March 2026 — drawing on a new Future of Privacy Forum (FPF) report examining seven African countries — documents how, in the absence of dedicated AI legislation, African policymakers have defaulted to existing data protection laws as a "backdoor" mechanism for AI governance, a pragmatic approach that analysts at FPF describe as a defining feature of Africa's second wave of digital policy reform but which increasingly shows structural limitations as AI moves from data processing into agentic decision-making, credit scoring, facial recognition, and digital lending. Mercy King'Ori, who leads FPF Africa from Nairobi, is candid about the limits: "There is a realisation that current data protection laws really don't cover all aspects of digital governance" — a gap that is now being addressed simultaneously in Nigeria (National Digital Economy and E-Governance Bill), Kenya (Artificial Intelligence Bill 2026), and South Africa (Draft National AI Policy), with the AU and AfCFTA Digital Trade Protocol adding a continental harmonisation layer. The piece frames the continent's regulatory moment as both urgent and structurally complex: standalone AI laws are finally in motion, but they arrive in countries where data protection enforcement institutions are still young, capacity is thin, and the primary tension between innovation and regulation is only beginning to be navigated.
A landmark report published by the Institute of Development Studies and the African Digital Rights Network reveals that African governments have spent at least $2 billion on Chinese-built AI-enabled smart city surveillance systems — including facial recognition, AI-powered CCTV, vehicle tracking, and command-and-control centres — across 11 countries: Algeria, Egypt, Kenya, Mauritius, Mozambique, Nigeria, Rwanda, Senegal, Uganda, Zambia, and Zimbabwe, with the total almost certainly higher because surveillance contracts are frequently secret and the study covers only 11 of Africa's 55 countries. Researchers found no compelling evidence that mass surveillance has reduced terrorism or serious crime in any of the countries studied — including Zambia and Senegal, which have no terrorist threat — while evidence that journalists, political opponents, and human rights activists are being tracked, arrested, and detained based on surveillance data is documented. Co-author Wairagala Wakabi of CIPESA warns that the "chilling effect" of unregulated mass surveillance is undermining democratic participation across the continent precisely as AI governance frameworks are being legislated — making the simultaneous legislative momentum in Nigeria, Kenya, and South Africa not merely a techno-regulatory matter but a democratic rights emergency.
With approximately 11 working days remaining before the National Assembly's end-of-March 2026 target, Nigeria's National Digital Economy and E-Governance Bill is the continent's most closely watched piece of AI legislation — the bill would grant NITDA authority to classify AI systems by risk, mandate annual impact assessments and operating licences for high-risk deployments in finance, public administration, surveillance, and automated decision-making, and impose fines of up to ₦10 million or 2% of annual Nigerian revenue for non-compliance. NITDA Director-General Kashifu Abdullahi has framed the legislation's purpose precisely: "You cannot be ahead of innovation, but regulation is not just about giving commands — it is about influencing market, economic and societal behaviour so people can build AI for good. That way, if there are bad actors, you can easily detect and contain them." Legal analysts at TechHive Advisory note that some experts now revise the timeline to Q2 2026 to account for possible amendments arising from the November 2025 public hearing, with five other active AI-related bills in the National Assembly creating a potentially complex legislative environment — but the convergence of Nigeria's bill, Kenya's Senate debate, and South Africa's imminent gazette represents the most concentrated legislative period in African AI history.
South Africa's Department of Communications and Digital Technologies has formally submitted the Draft National AI Policy to Cabinet for approval and gazetting, with legal bulletins from Fasken and Baker McKenzie — both actively circulating in South Africa's compliance and regulatory communities this week — confirming that gazetting for a 60-day public consultation period is expected to occur in March 2026 itself, before the final National AI Policy is submitted to Cabinet in 2026/2027 and sector-specific regulations follow in 2027/2028. The policy adopts a sector-specific, multi-regulator architecture — rejecting a single AI Act in favour of embedding AI governance within existing frameworks such as POPIA, the FSCA, and the Prudential Authority — structured around five pillars: skills capacity, responsible governance, ethical and inclusive AI, cultural preservation, and human-centred deployment. For every organisation deploying AI in South Africa's financial services, healthcare, or public-sector contexts, the unanimous advice from legal advisors is now immediate: the 60-day public comment window, once open, is the primary — and possibly only — opportunity to shape how sector-specific algorithmic explainability requirements, oversight mechanisms, and enforcement timelines are written before they become binding.
An iAfrica.com analysis published on 19 March 2026 argues that the convergence of AI-powered tools — from AI-assisted editing, visual effects, subtitling, and translation to concept development itself — with Africa's near-universal mobile penetration is creating a structural inflection point in the continent's creative economy, compressing the barrier between storytelling idea and finished film from expensive studio-based production to smartphone-and-AI pipeline accessible to any creator with a device and a story. The global creative economy is valued at more than $2 trillion and expanding rapidly, with Africa's share growing faster than any other region driven by a young and growing population and a global appetite for distinctly African narratives — and AI could accelerate this growth substantially by handling the technical production layer that previously required expensive training, equipment, and studio access, allowing African filmmakers, musicians, and digital creators to focus entirely on the storytelling itself. The significance for Africa's AI ecosystem extends beyond creative output: it positions mobile-first, creative-economy AI applications as potentially the continent's highest-reach use case, reaching demographics — youth creators, informal economy storytellers, rural artists — who will not be served by enterprise or fintech AI deployments.
Cassava Technologies — Africa's first and only NVIDIA Cloud Partner — deployed its AI Factory in South Africa on 18 March 2026, with a confirmed expansion roadmap to Nigeria, Kenya, Egypt, and Morocco, delivering GPU-as-a-Service and AI-as-a-Service through its CAIMEx multi-model exchange platform so that African enterprises, governments, and developers can build, fine-tune, and deploy AI solutions without routing sensitive data or workloads offshore. The AI Factory partners with the CSIR's National Integrated Cyberinfrastructure System and with Zindi — Africa's pan-continental data science competition platform with 70,000-plus developers — giving African researchers and builders access to local GPU resources for training models calibrated to African languages, datasets, and challenges rather than global defaults. As Group COO Ahmed El Beheiry stated at the launch: "For Cassava, building Africa's AI ecosystem is an act of empowerment, not just a technological milestone — we are ensuring that African businesses aren't just consumers of global tech, they are the architects of it," framing sovereign compute not merely as a commercial product but as the missing structural layer in every African national AI strategy to date.
A detailed Vellum Kenya legal analysis published on 19–20 March 2026 examines the structural tensions in the Artificial Intelligence Bill 2026 — noting that while the bill's intent to align Kenya with international standards, particularly the EU AI Act, is sound, its proposal to create three entirely new government bodies (AI Commissioner, AI Authority, AI Advisory Council) on top of the existing Office of the Data Protection Commissioner and Communications Authority risks fragmented oversight, regulatory confusion, and a compliance burden that will fall most heavily on local startups. The piece identifies what may be an irresolvable technical problem embedded in the bill's compliance architecture: many Kenyan developers do not build AI from scratch but adapt open-source global models for local use — models built and trained by companies in the US, Europe, or China — making it practically impossible for them to provide the audit trails and training-data transparency that the bill's high-risk classification system requires, since that information is not theirs to furnish. The analysis argues for a calibrated, risk-proportionate approach that combines Kenya's existing data governance infrastructure with adaptive principles rather than wholesale EU-style institutional replication — an argument with direct implications for every African country now designing first-generation AI legislation, from Nigeria to South Africa to Ghana.
Today's edition arrives at a genuinely historic intersection. Three distinct legislative processes — Kenya's AI Bill now before the Senate, Nigeria's end-of-March countdown, and South Africa's imminent gazette — are advancing simultaneously, meaning this month may produce more concrete AI regulatory infrastructure on the African continent than any preceding year. The IDS surveillance report (Rank #3) is a necessary corrective: it reminds us that the absence of legal frameworks has not meant the absence of AI deployment — it has meant the absence of accountability for systems that are already surveilling citizens at scale. The week's most important insight is that African AI governance is not a future consideration. It is a present emergency, and the builders, investors, and policymakers reading this newsletter are the people who will determine whether the frameworks being written this month protect people or merely regulate products.
Forward today's edition to a legal professional, a policymaker, or a developer building in Africa. Subscribe at simphiwemlotshwa.substack.com to receive every edition at 6:00 AM SAST.