AI Safety | Policy | Africa

Build it safe, or inherit it broken.

AI could add $2.9 trillion to Africa's GDP by 2030. But of the continent's 2,000+ languages, only 42 have any support in today's AI models. The African AI Safety & Policy Institute exists to close that gap — safely, skilfully, and on Africa's own terms.

Explore our work Partner with us

$2.9T Potential AI contribution to Africa's GDP by 2030

~5,000 AI safety professionals across the whole continent

42 / 2,000+ African languages with any support in today's AI models

Why now

AI is the century's defining infrastructure. Safety is Africa's defining choice.

Every century has its defining infrastructure — the railway, then electricity and the network. In ours, it is artificial intelligence: already the substrate beneath finance, medicine, agriculture, education, security, and speech. For a continent projected to be one in four people on Earth by mid-century, this infrastructure will shape more lives, sooner, than almost anywhere else.

The African Union's 2024 Continental AI Strategy set the ambition. AAISPI exists to build the institutional muscle behind it — closing the measurement, skills, and advocacy gaps so AI serves Africa, rather than being merely deployed upon it.

Mandate: Translate AI risk into practical, independent governance for African realities.
Audience: Governments, the African Union, funders, researchers, and civil society.
Standard: Evidence-led, Africa-tested, and independent of commercial or political capture.

The three gaps

Where safety research becomes public value.

01 Measurement Gap

No African country hosts a dedicated AI safety institute. Models are tested overwhelmingly in English — ask one to catch fraud in Hausa, misinformation in Akan, or harmful content in Swahili, and reliability collapses.

Independent benchmarks, red-teaming, and evaluation for African languages and contexts
Risk and harm assessments grounded in African threat models
Evidence that "safe" is tested, not assumed

02 Skills Gap

Africa's entire specialised AI talent pool is estimated at around 5,000 professionals — for a continent where 230 million jobs will demand digital skills by 2030.

A pipeline of researchers, auditors, and policy specialists
Training that makes safety a profession, not an import
Capacity every other AI ambition depends on

03 Advocacy & Representation Gap

National AI safety institutes now exist across the US, UK, EU, and beyond, writing the benchmarks the world inherits. Africa has had no permanent seat at that table.

Translating the AU Continental AI Strategy into concrete policy
An African voice where global AI standards are set — as an authority, not a guest
A home for the AU's proposed annual safety dialogue

What we do

Three gaps. One integrated mission.

Measurement

Independent AI safety benchmarks for African languages

Building red-teaming, evaluation, and risk assessment tools tested against African languages, contexts, and threat models — not assumed from performance elsewhere.

Skills

A talent pipeline for African AI safety

Cultivating the researchers, auditors, and policy specialists who make AI safety a profession on the continent, not a foreign import.

Policy

Translating the AU Continental AI Strategy into practice

Concrete, independent policy work that turns continental ambition into national capacity.

Policy

A home for Africa's annual AI safety dialogue

Convening governments, funders, researchers, and industry around the AU's proposed continental safety dialogue.

Timeline

From continental ambition to institutional muscle.

2024

African Union

Continental AI Strategy adopted

One of the most ambitious regional AI frameworks anywhere, setting the ambition AAISPI exists to make real.
2025

Research consensus

The case for an African AI safety institute is documented

Researchers across the field reach the same conclusion: African perspectives are largely absent from global AI safety, with no dedicated institute to represent them.
2026

Implementation

The Continental AI Strategy enters its first implementation phase

The blueprint exists — agile safety regulation, an annual continental safety dialogue, a centre of excellence. What has been missing is the independent institution to carry it.
2026

Founding

The African AI Safety & Policy Institute is founded

AAISPI convenes as the independent vehicle for measurement, skills, and advocacy — Africa's seat at the table it helped design.

Convening power

A trusted table for AI safety, governance, and public interest.

AAISPI creates space for governments, universities, civil society, funders, and industry partners to move from abstract concern to shared standards and workable policy — including hosting the annual continental AI safety dialogue proposed in the AU's own strategy.

Policy roundtables for senior decision-makers
Evidence briefings for public institutions
Cross-sector research partnerships

Partner with AAISPI