Introducing Africa's AI Almanac (Previously Liminal AI)

Welcome to the inaugural edition of Africa's AI Almanac.

We are launching this monthly briefing at a pivotal moment. As AI transitions from a technical curiosity to a foundational pillar of African economies, the rules of the game are being written in real time.

This newsletter is designed to move beyond a “list of links.”

Every month, we will bring you a cohesive narrative that connects the dots between global tech shifts and local regulatory realities: the policy breakthroughs, the technical shifts, and the deep-seated ethical challenges that define our path forward. Our objective is to move beyond reactionary reporting. We will explain why developments matter, how they intersect, and what they signal for the continent’s AI future. Our goal is to ensure you are not just informed, but equipped to shape an AI future that is safe, sovereign, and uniquely African. Thank you for taking this journey with us.

2025 - The retrospective

The narrative on AI in Africa shifted decisively in 2025, becoming impossible to ignore. The year began with a push for continental consensus at the inaugural Global AI Summit on Africa in Kigali, where countries adopted the Africa Declaration on AI. This commitment to harmonising national frameworks supports the newly constituted Africa AI Council, signalling a clear intent to avoid regulatory fragmentation. This momentum aligns with the Continental AI Strategy, which is now one year into its five-year implementation timeline.

The continent is no longer a “rule-taker” in global tech diplomacy. African countries have taken their seats at the global table, with Kenya, Nigeria, and Rwanda participating in the Seoul Declarations on AI safety. Meanwhile, South Africa has leveraged its BRICS membership and G20 Presidency to strengthen cooperation among the Global South on AI policy.

Besides this high-level consensus, countries are pursuing a dual track of innovation-first strategies and risk-mitigating regulation:

• Strategic Frameworks: Côte d’Ivoire, Kenya, Libya, and Nigeria adopted national AI strategies this year. These frameworks provide the infrastructure necessary to balance economic growth with ethical safeguards.

• The Move to Binding Law: Angola published a robust draft AI Law that prioritises transparency and human intervention. Notably punitive, it stipulates sanctions ranging from heavy fines to prison terms of up to 12 years. Nigeria has introduced a complex suite of bills, including one that positions the National Information Technology Development Agency as a “super-regulator” with powers to regulate AI.

• The “Agency” Model: Ghana’s Emerging Technologies Bill adopts a cross-functional approach, establishing an agency to enforce compliance and mandating a 50% cap on non-local equity for tech businesses. Mauritius established a dedicated AI Unit, and Chad appointed a head to lead its newly created Directorate General of Artificial Intelligence.

With multiple bills advancing in parliaments across Morocco and Namibia, and with more countries indicating interest in introducing an AI-specific law, the race to enact the continent’s first dedicated AI law is underway. We expect to see the first “hard-law” codification before Q2 of 2026. This transition provides the necessary “teeth” for enforcement but also signals a shift toward codified mandates that companies must begin to operationalise now to avoid costly retrofitting later.

Advancements in AI from around the world that have had a ripple effect in Africa include the release of DeepSeek. DeepSeek’s emergence as a powerful, low-cost large language model was a significant moment for Africa, not because it is “another AI model,” but because it challenged who gets to build advanced AI and on what terms. By demonstrating that high-performing models can be trained with fewer resources, DeepSeek challenged the assumption that frontier AI is the exclusive domain of well-capitalised firms. For African countries and institutions, this matters.

DeepSeek reopened the conversation on AI sovereignty. If capable models can be trained or adapted with lower compute costs, African universities, startups, and public institutions gain a more realistic pathway to local AI development. This does not mean training frontier models from scratch everywhere, but it does mean greater scope for regional fine-tuning, language support, and context-specific applications, areas where Africa has long been underserved.

​​Besides the excitement, DeepSeek highlighted the policy trade-offs Africa must confront. Lowering barriers to AI development increases opportunities for innovation but also raises governance questions, including data sourcing, model transparency, security risks, and alignment with local values. African AI strategies that focus only on adoption, without addressing accountability, data governance, and capacity building, risk repeating earlier patterns of technological dependency

The greatest conclusion being that while DeepSeek is not Africa’s solution, it is a signal. The window for Africa to move from passive AI consumer to active shaper is narrow, and it is as much a governance challenge as a technical one.

To bridge the AI development gap, notable partnerships aimed at increasing investments in AI infrastructure and adoption around the continent marked the year’s developments. Some notable ones were:

1. Launch of the African Hub on AI Safety, Peace & Security

In September 2025, the University of Cape Town and the Global Centre on AI Governance launched the African Hub on AI Safety, Peace and Security. This institute is designed to centre African perspectives in global AI safety discussions, advance research on safe AI systems, build continental capacity, and influence policy frameworks that reflect African priorities in AI governance and security

2. ALX, Anthropic & Government of Rwanda - AI Learning Initiative

A landmark partnership was announced between ALX, Anthropic, and the Government of Rwanda to deploy an AI-powered learning assistant called Chidi. This collaboration aims to modernise education and position African learners not just as users of AI but as future creators and innovators within the AI economy.

3. G7-Endorsed AI Hub for Sustainable Development Partnerships

Under a G7-backed initiative, a collaboration was launched to empower African innovators in AI for sustainable development. Partners included Zindi, AfriLabs, and OpenStartup International, working through an AI Hub platform that provides mentorship, compute resources, and global networks to local developers solving regional challenges.

Undeniably, Africa is entering the AI era at a moment when the global governance playbook is still being written. This creates leverage. Decisions taken today on institutions, laws, infrastructure, and partnerships will determine whether AI deepens existing asymmetries or becomes a tool for shared growth, public value, and continental resilience.

What is clear is that activity is no longer speculative. From continental declarations and national strategies to draft laws, safety institutes, and infrastructure partnerships, AI governance is rapidly becoming operational. The question is no longer whether Africa will regulate AI, but how, by whom, and in whose interests.

In the editions ahead, the newsletter will follow these questions closely, cutting through noise to surface what truly matters for policymakers and practitioners. As AI continues to reshape economies and power structures, staying ahead of governance choices will be as important as keeping pace with the technology itself.

This is a shared journey. We want your feedback on what hit the mark and which complex issues you would like us to look into next. If this briefing proved valuable, please share it with a colleague or partner working at the intersection of AI and policy.

We are glad to have you with us at the start.

Until next month,

The Editorial Team