Skip to content
Apart Research
Apart Research
Contact us
Sprint projectJun 21, 2026Nairobi, Kenya

Africa AI Risk Index (AARI)

Ray Munene · Team Solve-it

Submitted to Global South AI Safety Hackathon. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Africa AI Risk Index (AARI)

Open in a new tab (opens in new tab)
Code (opens in new tab)
Share

The Africa AI Dependency Risk Index is a strategic evaluation framework that measures an organization's exposure to foreign AI technologies by scoring 34 variables across six weighted pillars: Model Dependency, Cloud Dependency, Data Sovereignty, Local Ecosystems, Compute Access, and Governance. Using a weighted risk-conservative algorithm, it aggregates these dimensions into a single index score ($0-100%$) and outputs targeted mitigation policies ranging from standard API diversification to the immediate establishment of a Sovereign AI Council and local workload migration, to safeguard digital autonomy and infrastructure resilience.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

Only the Apart team reads this, and funders if you agree below.

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical AI safety problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.

  1. The project definitely has potential and offers an interesting perspective on how to think of the different aspects that AI sovereignty may be composed of. I would love to see an improved and more developed as well as nuanced version of this index.

    Directions for further exploration and improvement:

    (1) It is unclear currently how an organization scores a specific number - how to quantify an evaluative statement?

    (2) There is more nuance to AI sovereignty than the current pillars present. E.g. they are not all necessarily independent pillars but some are dependent on each other - how do we score them then?

    (3) Clarify the audience more - this will also determine how you present the information and what you highlight.

    (4) The section on strategic actions seems ad hoc and not realistic. Would recommend consulting with local stakeholders on this to identify also any existing pathways to build on or learn from practices in other fields/jurisdictions.

    (5) Add a limitations section about your own methodology and approach, to also allow others to build on it or reproduce.

    Read full reviewShow less

  2. The project could benefit from consideration of two key considerations. The first is under the governance and compliance section where it would be interesting to score based on responsible AI efforts as opposed to policy existence, audit rates, incident runbooks, etc. Under this section the project could, for example, consider an index such as the Global Index on Responsible AI, which highlights that policy alone is not an adequate assessment for responsible / safe / ethical design, development, and deployment of AI. The second consideration that would have been useful to interrogate is that digital sovereignty could also serve an exploitative function (e.g. if an autocratic government has access to all citizen data).

  3. The project is well-executed and addresses an important problem. However, there are two notable limitations:

    - The index is presented as a scoring architecture (six dimensions, 34 variables, specified weights, risk bands) but is never instantiated against a real institution or country. No case study assigns scores, no sensitivity analysis is performed on the 20/18/18/16/14/14 dimension weights, and the vendor-diversity formula (R_vendor = max(10, 100 − 18×(V−1))) is asserted without evidence that its curve reflects observed dependency risk. As presented, the framework cannot be validated, compared to alternative weightings, or falsified.

    - The variable list is broad but definitions are shallow: "workload API percentage," "multi-cloud maturity," or "R&D investments" require operational specifications and data sources to become measurable, and none are provided. Because the "conservative default = highest risk" rule pushes any missing input to 100, an organization with even moderate data gaps would receive a near-critical score regardless of its actual posture — meaning the scores the framework produces will be driven by data availability more than by dependency risk, which undercuts the diagnostic purpose.

    Read full reviewShow less

Cite this project

@misc{munene2026africa,
  title = {{Africa AI Risk Index (AARI)}},
  author = {Ray Munene},
  year = {2026},
  month = jun,
  note = {Submitted to Global South AI Safety Hackathon, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/africa-ai-risk-index-aari-vrpy}},
  url = {https://apartresearch.com/sprints/projects/africa-ai-risk-index-aari-vrpy}
}

More from Global South AI Safety Hackathon

All 216 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026