leviupendo/africa-ai-sovereignty-scorecard

# The Sovereignty Index — by Lucy Jong'a A tracker for whether African nations are becoming co-architects of the AI era — or just a consumer market for it. ## What it tracks Every country is scored 0-3 across three pillars, based on published national AI strategies and data protection law: | Pillar | What it measures | |---|---| | **Data governance** | Legal protections on how domestic data is collected, stored, and exported | | **Local capacity building** | Investment in local AI training facilities, research institutions, and skills programs | | **Firm accountability** | Requirements for foreign AI companies to partner with, employ, or share value with local institutions | Scores are summed into a 0-100 sovereignty score per country. ## Coverage Kenya, Nigeria, South Africa, Rwanda, Egypt, Ghana, Morocco, Tunisia, Senegal — 9 countries, every score sourced to a primary document (see `data.js`). ## Sourcing and review flags Every score carries a `source`, a one-line `rationale`, and a `confidence` level: - **`high`** — well-evidenced by a primary source (an enacted law, a documented enforcement action, a funded and named program). - **`review`** — a judgment call. The underlying fact is sourced, but translating it into a 0-3 score involved a subjective weighting worth a domain expert's scrutiny. As of the last research pass (July 2026), 11 of 27 pillar scores are flagged `review`. ## Running locally No build step, no dependencies. **Option A — just open it** Double-click `index.html`, or open it directly in a browser. **Option B — serve it (recommended)** ```bash python3 -m http.server 8000 ``` Then visit `localhost`. ## Project structure ``` . ├── index.html # page structure, branded for Lucy Jong'a ├── style.css # dark, finance/strategy-toned styling ├── data.js # the sourced policy data + scores ├── app.js # renders the scorecard from data.js └── README.md ``` ## Updating scores All scoring lives in `dat …