The African Union Continental AI Strategy: What It Says, What It Has Done and What It Still Misses
Abstract
The African Union Continental Artificial Intelligence Strategy, adopted
by the AU Executive Council in Accra, Ghana in July 2024, represents the most
comprehensive AI governance document produced at the continental level in
Africa.1
Two years into its existence, this article conducts a systematic policy
analysis of the Strategy. It examines what the document says across its six
focus areas and six implementation pillars, reviews what has been achieved
under Phase 1 of its implementation roadmap (2025 to 2026), and identifies ten
substantive gaps where the Strategy falls short of what effective continental
AI governance requires. The analysis draws on the Strategy's own text, the 2025
Kigali Africa Declaration on AI, implementation data from academic and
institutional sources and national policy developments across the continent.
The article is intended to contribute to the ongoing policy dialogue on
how the AU's continental framework can be strengthened as it moves from its
foundational phase toward Phase 2 implementation from 2028 onwards.
1. Introduction
When the African Union Executive Council endorsed the Continental
Artificial Intelligence Strategy on 18 and 19 July 2024, it was widely
described as a historic step.1 For the first time, Africa's 55 member
states were bound together by a shared continental framework on artificial
intelligence. The Strategy aligned with Agenda 2063, the AU's 50-year blueprint
for continental transformation, and with the United Nations Sustainable
Development Goals.2
The adoption followed an extensive consultative process involving member
states, development partners and international organisations. UNESCO played a
notable role in shaping the ethical frameworks contained in the document.3 The
Strategy was positioned not as a regulatory instrument but as a guiding
framework, one that would inform national AI strategies, continental governance
structures and investment priorities.
Two years into its life, it is appropriate to subject the Strategy to
rigorous analysis. Policy documents of this scale require assessment not only
at the level of their stated intentions but also at the level of what they
leave out, what they underspecify and what progress has actually been made
against their own targets. This article undertakes that assessment.
2. What the AU Continental AI Strategy Says
The Strategy is structured around six focus areas, which it describes as
pillars of continental action, and six implementation pillars that define how
the strategic objectives are to be pursued.
2.1 The Six Focus Areas
The first focus area is maximising the benefits of AI for social
and economic development and cultural renaissance. The Strategy frames AI as a
transformative tool for healthcare, agriculture, education, finance and public
service delivery.4
The second focus area is minimising risks and safeguarding AI
development from harm. The Strategy acknowledges risks including algorithmic
bias, potential discrimination against women and vulnerable persons,
cybersecurity threats, privacy violations and the spread of misinformation.4
The third focus area is building capabilities in infrastructure,
datasets, computing, skills and education, research and innovation, and
specialised AI platforms. The Strategy acknowledges that sustained investment
in reliable electricity, broadband connectivity, data infrastructure and
computing power is a prerequisite for any meaningful AI development on the
continent.4
The fourth focus area is fostering regional and international
cooperation. The Strategy calls for partnerships between African states and
between Africa and global actors, with the goal of building the continent's
capacity to leverage AI in ways that promote social and economic development.1
The fifth focus area is accelerating AI investment. The Strategy
calls for stimulating public and private investment at national and regional
levels, referencing blended finance models and public-private partnerships as
the primary instruments.4
The sixth focus area is creating an inclusive governance and
regulatory framework. The Strategy calls for agile, forward-looking, risk-based
regulations at national and regional levels, alignment with existing
continental frameworks and the adoption of governance principles rooted in
human rights, Ubuntu values and African cultural heritage.5
2.2 The Six Implementation Pillars
Alongside the focus areas, the Strategy identifies six implementation
pillars: human capital development for AI; infrastructure and data foundations;
an enabling environment for AI; the AI economy; sustainable partnerships; and
monitoring and evaluation.6 These pillars are designed to translate the
strategy's focus areas into operational activities.
The monitoring and evaluation pillar is notable in that it explicitly
frames the Strategy as a living document, subject to periodic review and
update.6
This acknowledgement of adaptability is one of the Strategy's more credible
institutional design features.
2.3 The Implementation Roadmap
The Strategy sets 15 strategic objectives to be achieved by 2030,
organised across a two-phase implementation roadmap.7 Phase 1,
running from 2025 to 2026, focuses on establishing governance frameworks,
developing strategic documents, mobilising resources and building capacities.
Phase 2, commencing from 2028, concentrates on executing core projects at
scale.
The eight principles that guide the Strategy are: AI should address local
African needs first; AI should be people-centred; AI should uphold human rights
and dignity; AI should advance peace and security; AI development and use
should be inclusive and respect African diversity; AI should be ethical and
transparent; governance should foster regional cooperation; and AI solutions
should be supported by formal and informal education.7
The AU Strategy is
broad in its principles and ambitious in its targets. Its stated ambition is to
position Africa as a global leader in ethical, responsible and equitable AI.
The question is whether the architecture behind those principles is sufficient
to deliver that outcome.
3. What Has Happened Since July 2024: A
Progress Review
The period from July 2024 to July 2026 corresponds to Phase 1 of the
Strategy's implementation roadmap. Several significant developments have
occurred, though progress has been uneven.8
3.1 The Kigali AI Declaration and the $60 Billion Fund
The most significant continental AI development since the Strategy's
adoption was the signing of the Africa Declaration on Artificial Intelligence
at the inaugural Global AI Summit on Africa in Kigali, Rwanda, on 4 April 2025.9
Forty-nine African countries endorsed the Declaration, which outlines seven
priority pillars: talent, data, infrastructure, markets, investment, governance
and institutional cooperation.
The Declaration announced a USD 60 billion Africa AI Fund, described as a
financing mechanism combining public, private and philanthropic capital to
support infrastructure, institutions and research.10 It also established the
Africa AI Council, co-chaired by the African Union Commission and the
International Telecommunication Union and housed under the Smart Africa
Steering Committee, to coordinate governance and policy implementation across
public, private, academic and civil society actors.11
A notable private sector commitment accompanied the Declaration: Cassava
Technologies announced an investment of up to USD 720 million in partnership
with Nvidia to build the first Africa AI Factory, providing AI infrastructure
to Egypt, Kenya, Morocco, Nigeria and South Africa.8
3.2 High-Level Policy Dialogue and Continental Momentum
In May 2025, the AU Commission convened a High-Level Policy Dialogue on
the Development and Regulation of AI in Africa in Addis Ababa, drawing
participants from over 40 African countries. The dialogue produced a Communique
declaring AI a strategic priority for the continent.12 It
committed to developing the necessary digital infrastructure, high-quality
datasets and compute capabilities to underpin responsible AI development.
National-level momentum has also accelerated. In the first quarter of
2025, Ivory Coast, Kenya and Namibia published national strategies. Lesotho and
Tanzania released draft strategies. By mid-2026, at least 15 African countries
have adopted official AI frameworks, with several others in active development.8
3.3 Funding Reality vs. Announcement
The gap between announced ambition and confirmed funding is the starkest
measurable finding from Phase 1. An 18-month implementation analysis covering
July 2024 to October 2025 found that confirmed commitments represented only
0.26 to 0.60% of estimated continental AI investment requirements.13
Continental AI needs are estimated at USD 500 billion. More than USD 7 billion
is required specifically to address deficiencies in data, computing and skills.8
The same analysis found that 83% of AI startup funding in the first
quarter of 2025 went to Kenya, Nigeria, South Africa and Egypt.12 This
geographic concentration reflects a structural inequality within the continent
that the Strategy acknowledges but has not yet addressed through concrete
redistribution mechanisms.
Confirmed funding
commitments under Phase 1 represent less than 1% of estimated continental AI
investment needs. The USD 60 billion Africa AI Fund remains an announcement
without a published disbursement mechanism.
4. What the AU Strategy Still Misses: Ten
Critical Gaps
The scorecard table below maps each identified gap against the Strategy's
stated positions. The sections that follow examine each gap in depth.
Table 1: AU Continental AI Strategy —
Policy Scorecard (July 2026)
|
Policy Area |
What the AU
Says |
What It
Misses |
|
Funding |
Calls for public-private
investment and blended finance |
No binding commitment
levels; confirmed funds = 0.26-0.60% of estimated needs |
|
Data Sovereignty |
Promotes National Data
Strategies and data localisation |
Over 80% of African cloud
services remain foreign-controlled; no enforcement timeline |
|
Enforcement |
Multi-level governance model
proposed |
No binding mechanisms;
strategy remains non-binding on member states |
|
Language Inclusion |
Recognises linguistic
diversity as an AI risk area |
No specific multilingual AI
standards or requirements; 2,000+ African languages unaddressed |
|
Gender Equity |
Calls for inclusion of women
and marginalised groups |
No measurable gender
targets; 86% of African women lack basic AI proficiency |
|
Civil Society |
Mentions multi-stakeholder
consultation |
No formal civil society
representation structure in continental governance bodies |
|
Small States |
Aims for continent-wide
participation |
83% of AI startup funding
concentrated in four countries; no redistribution mechanism |
|
Implementation |
Two-phase roadmap
(2025-2026, 2028+) |
Phase 1 governance
structures largely aspirational; Africa has only 3% of global AI talent |
|
Safety and Risks |
Addresses bias, privacy and
misinformation |
No provisions for autonomous
weapons, advanced AI safety or catastrophic risk scenarios |
|
AfCFTA Alignment |
References free trade area
for AI market development |
No binding digital trade
harmonisation mechanism between member states as yet |
Source:
Author's analysis based on AU Continental AI Strategy (2024), Kigali
Declaration (2025) and implementation review sources cited below.
4.1 Gap One: Funding Mechanisms Lack Binding Commitments
The Strategy calls for blended finance, public-private partnerships and
engagement with development partners including UNESCO, the IDRC, SIDA, the
World Bank and the African Development Bank.13 However, it does not specify
minimum national budget allocations for AI, establish binding contribution
levels for member states or provide a disbursement architecture for the USD 60
billion Africa AI Fund announced in Kigali.
Without enforceable financial commitments, the Strategy's investment
pillar depends entirely on voluntary action from governments that already face
competing fiscal pressures. The result, as Phase 1 data confirms, is a
significant gap between announced ambition and deployed capital.
4.2 Gap Two: Data Sovereignty Lacks an Enforcement Timeline
The Strategy promotes the development of National Data Strategies and
data localisation frameworks, and acknowledges the importance of data
sovereignty for African AI development.5 However, as of mid-2026, over
80% of Africa's cloud services remain under the control of foreign entities.14 The
Strategy provides no enforcement timeline, no minimum data localisation
standard and no penalty mechanism for member states that fail to advance data
sovereignty.
The risk identified by multiple observers is that global foundational AI
models such as those produced by major US and European technology companies
rely almost entirely on data collected, processed and stored outside Africa.
This pattern, if unaddressed by policy, reproduces historical patterns of
resource extraction through what some analysts describe as digital colonialism.14
4.3 Gap Three: No Binding Enforcement Mechanisms
The Strategy proposes a multi-level governance model involving the AU
Commission, regional economic communities, member states and the private
sector.4
However, the Strategy itself is non-binding. Member states face no legal
obligation to implement its provisions, adopt its principles or report on their
progress against its objectives.
Assessment of Phase 1 progress finds that governance structures proposed
in the Strategy remain largely aspirational.13 Observers describe the Kigali
Declaration and the Continental Strategy as at risk of remaining ink on paper
due to absent enforcement mechanisms and weak coordination among politically
unstable member states.15
4.4 Gap Four: Linguistic Diversity Is Recognised but Not Operationalised
The Strategy acknowledges Africa's cultural and linguistic diversity as a
guiding principle and identifies linguistic barriers as one of five central
gaps impeding AI adoption.4 However, it does not specify requirements for
multilingual AI development, set targets for language inclusion in AI systems
or establish standards for AI tools to operate across Africa's more than 2,000
languages.
Ethiopia's national AI policy explicitly prioritises indigenous large
language model development as a matter of linguistic sovereignty. Rwanda's
policy includes provisions for linguistic diversity in AI applications. The
continental strategy, by contrast, treats linguistic inclusion as a background
principle rather than an operational requirement.16
— ADVERTISEMENT
—
Are you ready to work, earn and grow with AI?
Fluent in AI: The Practical Guide for Africans Who Want to Work, Earn
and Grow with Artificial Intelligence
Written for African professionals, students and entrepreneurs.
Practical, actionable and jargon-free.
Get your free copy: https://selar.com/611u156116
4.5 Gap Five: Gender Equity Lacks Measurable Targets
The Strategy calls for AI to empower African women and address the
specific needs of marginalised groups. It positions gender inclusion as a
guiding principle.4 However, it sets no measurable targets for
women's participation in AI development, governance or training. This omission
is significant given the scale of the problem it is silent on.
An ImpactHER 2024 survey across 52 African countries found that 86% of
women lack basic AI proficiency, 60% have not received digital skills training
and 50.2% lack quality internet access, a gap 37 percentage points below the
rate for African men.13 Women represent only 15 to 20% of participants
in most technical AI training programmes on the continent. Without binding
gender targets, the Strategy's inclusion principles remain aspirational.
4.6 Gap Six: Civil Society Has No Formal Governance Role
The Strategy references multi-stakeholder consultation and notes that
civil society was involved in the document's development. However, it does not
create a formal civil society representation structure within the continental
governance framework.5
The Africa African Governance Forum (AAGF) has emerged as an independent
platform connecting governments, the private sector, civil society and
academia. Its emergence reflects a gap that the Strategy's own governance
architecture does not adequately fill.17 Effective AI governance
requires civil society not only to be consulted but to have a structured,
ongoing role in oversight and accountability.
4.7 Gap Seven: Small and Fragile States Are Left Behind
The Strategy commits to leaving no African state behind. In practice, the
concentration of AI investment and talent in a small number of countries is
deepening intra-continental inequality rather than narrowing it.12 Africa
accounts for only 3% of the global AI talent pool. Fewer than 25,000
specialised graduates are produced annually against an estimated continental
need of 300,000.15
The Strategy does not include a redistribution mechanism, a tiered
implementation framework or differentiated support for small, fragile or
conflict-affected states. Without these, continental AI governance will remain
concentrated in its strongest economies and the gap between them and the rest
will widen.
4.8 Gap Eight: AI Safety and Catastrophic Risk Provisions Are Absent
The Strategy addresses conventional AI risks including bias, privacy
violations, cybersecurity threats and misinformation. These are legitimate and
important areas of concern.4 However, the document contains no provisions
regarding autonomous weapons systems, advanced AI safety concerns, frontier
model risk or catastrophic risk scenarios.18
As global AI capability advances rapidly and the geopolitical contest
over AI dominance intensifies, continental governance frameworks that do not
address these dimensions will be structurally incomplete. African citizens and
institutions are exposed to these risks regardless of where the technologies
are developed.
4.9 Gap Nine: AfCFTA Integration Remains Aspirational
The Strategy references the African Continental Free Trade Area as a
platform for AI market development and cross-border digital trade.2 However,
digital services trade on the continent remains highly uneven and regulatory
fragmentation between member states continues to constrain cross-border AI
deployment.19
The absence of a binding digital trade harmonisation mechanism between
the Strategy and the AfCFTA framework means that the potential of the
continental market for AI cannot be effectively mobilised. African AI startups
operating across borders face different regulatory regimes in virtually every
jurisdiction they enter.
4.10 Gap Ten: Implementation Science Is Underdeveloped
The Strategy's monitoring and evaluation pillar acknowledges the need for
ongoing review. However, the document does not establish a clear methodology
for measuring implementation progress, define indicators for each strategic
objective or specify who is responsible for collecting and publishing that
data.6
As of mid-2026, there is no publicly available AU progress report on
Phase 1 implementation against the Strategy's 15 strategic objectives. The
18-month implementation analysis referenced in this article was produced by
independent researchers, not the AU Commission.13 Without transparent,
publicly accessible progress reporting, accountability for the Strategy's
ambitions cannot be established.
A continental AI
strategy without a public implementation tracker is a strategy without
accountability. The AU Commission should publish an annual implementation
report against each of the 15 strategic objectives.
5. What This Means for You
The AU Continental AI Strategy affects every African citizen,
professional, student and institution, whether or not they are aware of its
existence. Understanding what it does and does not do enables more informed
participation in the decisions that will shape AI governance on the continent.
For policymakers and government officials,
the gaps identified in this article represent an agenda for Phase 2
preparation. Binding financial commitments, a disbursement architecture for the
Africa AI Fund, measurable gender targets and a formal civil society governance
role are all achievable policy additions that would substantially strengthen
the Strategy's credibility.
For researchers and academics,
the absence of public implementation data is both a gap and an opportunity.
Independent monitoring of Phase 1 progress, rigorous assessment of national
strategy alignment with the continental framework and analysis of the Africa AI
Fund's disbursement process as it develops are areas where African scholarship
can contribute directly to better governance.
For civil society organisations,
the absence of formal civil society representation in the AU's continental AI
governance architecture is a structural gap that requires active advocacy.
Civil society engagement in national public consultations, such as South
Africa's ongoing draft policy comment process, is one mechanism for ensuring
citizen voices reach the continental level.
For entrepreneurs and developers,
the Africa AI Fund, once operationalised, and the regulatory sandbox provisions
in the Kigali Declaration represent potential access points for investment,
particularly for those working in underserved regions or on applications in
healthcare, agriculture, education and local languages.
For students and early-career
professionals, the Strategy's human capital development pillar and
the Kigali Declaration's commitment to continent-wide curriculum development
signal that AI skills will attract sustained public investment across the
continent. Positioning yourself now, through AI literacy, data skills and
policy engagement, places you at the centre of where Africa is directing its
institutional energy.
6. Conclusion
The AU Continental AI Strategy is a significant and genuinely
consequential document. It established the first continent-wide AI governance
framework for Africa, anchored in African values, aligned with long-term
development objectives and designed to be adaptive over time.1 The
momentum it generated contributed directly to the Kigali Declaration, the
Africa AI Fund announcement and the establishment of the Africa AI Council.
At the same time, a rigorous assessment of what the Strategy says, what
has been achieved in Phase 1 and what it still misses reveals a document that
is strong on principles and weak on mechanisms. Its funding commitments are
voluntary. Its enforcement architecture is non-binding. Its data sovereignty
provisions lack timelines. Its gender targets are absent. Its linguistic
inclusion provisions are underspecified. Its civil society governance role is
informal. Its implementation progress is unmeasured in public.
None of these gaps are insurmountable. They are design choices that can
be revised. The movement from Phase 1 to Phase 2, from 2028 onwards, provides a
critical window to address them. African policymakers, researchers, civil
society organisations and practitioners who engage with the Strategy's
shortcomings now will be better positioned to influence how Phase 2 is
designed.
The AU Continental AI Strategy is the right foundation. Building a
framework that genuinely delivers on its ambitions requires more than
endorsement. It requires accountability, resources and the institutional will
to translate principles into enforceable action.
References
1.
African Union. (2024). Continental
Artificial Intelligence Strategy. AU Executive Council 45th Ordinary Session,
Accra, July 18-19, 2024. au.int
2.
African Union. (2024). Continental
Artificial Intelligence Strategy: Executive Summary.
au.int/sites/default/files/documents/44004-doc-EN-_Continental_AI_Strategy_July_2024.pdf
3.
Future Center for Advanced Research
and Studies. (2025, December). Ambition or Inability: Assessing the African
Union's Efforts to Establish AI Governance. futureuae.com
4.
Ilina Program. (2024). The AU
Continental AI Strategy: Concrete Safety Proposals or High-Tech Hype?
ilinaprogram.org
5.
Future Privacy Forum. (2024, November
18). The African Union's Continental AI Strategy: Data Protection and
Governance Laws. fpf.org
6.
YesChat.AI. (2024). The African Union
AI Continental Strategy: Examining the African AI Landscape. yeschat.ai
7.
Young African Policy Research Hub.
(2024). African Union Continental AI Strategy: A Vision for Africa's
Technological Future. youngafricanpolicyresearch.org
8.
Carnegie Endowment for International
Peace. (2025, September). Understanding Africa's AI Governance Landscape:
Insights from Policy Practice and Dialogue. carnegieendowment.org
9.
Africa Declaration on Artificial
Intelligence. (2025, April 4). Kigali, Rwanda. aiexpoafrica.com
10.
BABL AI. (2025). African Leaders Sign
Landmark AI Declaration in Kigali, Pledge USD 60 Billion Fund. babl.ai
11.
Startup Researcher. (2025). Africa
Unveils AI Declaration, USD 60B Fund for Innovation. startupresearcher.com
12.
African Union Commission. (2025, May).
Africa Declares AI a Strategic Priority. au.int
13.
ResearchGate. (2025, October). From
Aspiration to Implementation: The AU AI Continental Strategy and the
Development-Governance Paradox. researchgate.net
14.
AI Reports Africa. (2025, September).
Africa Accelerates AI Governance: Experts Weigh In on Policy, Sovereignty and
Innovation. aireports. Africa
15.
Bennett School of Public Policy,
University of Cambridge. (2025, September). Strategic paradox: how the African
Union performs the future through AI policy. bennettschool.cam.ac.uk
16.
Oxford Academic / Science and Public
Policy. (2026, April). Strategy as governance: the governance of AI in Africa.
academic.oup.com/spp
17.
African Business. (2026, March 22).
Africa stakes its claim in global AI governance. African. Business
18.
Oxford Academic / Oxford Intersections
AI in Society. (2025, March). Africa's Continental AI Strategy: A Black
Mirror's Mirage or a Fast Track for Development? academic.oup.com
19.
CIPESA. (2026, February). Why Data and
AI Governance Are Central to Africa's Digital Trade Ambitions. cipesa.org
20.
Startup Graveyard Africa. (2025).
Decoding the AU Continental AI Strategy. startup graveyard. Africa
Comments
Post a Comment