Jonas Kgomo · Founder, Equiano Institute

AI governance & safety

Grounding AI governance in African contexts and empirical safety research.

I’m an engineer and founder of Equiano Institute, working on AI governance and technology policy in Africa. My work connects policy research with questions about how AI systems behave, whom they serve, and how we can assess their risks. Here is a short introduction through selected public work.

Explore the research

A record you can examine

Three co-authored contributions. Each links to the paper, its authors and its findings.

Policy & institutions

2024 · Co-author · Published in Data & Policy, 2025

AI policy in African contexts

Case studies of four African nations examine what global AI readiness assessments miss about local capabilities and policy development.

Governance relevanceA foundation for policy assessments that reflect the institutions and communities they serve.

Language & reliability

2024 · Co-author · Working paper

Testing truthfulness across languages

Uhura evaluates scientific question answering and truthfulness in six African languages. The models tested performed better in English, exposing gaps in multilingual reliability.

Governance relevanceEvidence for evaluating safety in the languages people actually use.

Persuasion & agency

2025 · Co-author · Preprint, revised 2026

Measuring persuasive capabilities

Experiments compare LLMs with incentivised human persuaders in a conversational quiz. The revised study finds that persuasive advantages depend on the model and whether persuasion is truthful or deceptive.

Governance relevanceA way to study manipulation risks while keeping claims tied to the conditions tested.

From evidence to useful institutions

Equiano is the institutional thread running through my focus on African AI governance. Alongside this, my public work includes Explore for AI policy research and work with Fairwork on AI labour governance. The aim is to connect technical evidence with decisions about accountability, participation and the distribution of AI’s benefits.

For collaborators & funders

Help shape the next piece of work.

I welcome conversations about joint research, multilingual evaluations and support for African AI governance. A useful starting point is a specific question, a shared research output, or a funding priority we could develop together.