SAAIL Lab SAAIL Lab
AAAI 2026 Workshop (EGSAI) 1+ Citations Published · AAAI 2026 Workshop

AI Sovereignty in the Global South: Power, Dependency, and Strategic Futures

S. Davies 5 , Vikranth H. Nagaraja 6 , Innocent Nyalala 1,2 , Nirav Bhatt 1 , Soumya Banerjee 5
1School of Engineering and Science, Indian Institute of Technology Madras Zanzibar, Tanzania
2Wadhwani School of Data Science and AI (WSDSAI), Indian Institute of Technology Madras, Chennai, India
3College of Engineering, Nanjing Agricultural University, Nanjing, China
4School of Information Technology, Deakin University, Australia
5Department of Computer Science and Technology, University of Cambridge, UK
6School of Science, Engineering and Environment, University of Salford, UK
AAAI 2026 Workshop on Equitable and Sustainable Global AI (EGSAI), Philadelphia, USA (Jan 2026)

Summary: Analyzes compute, talent, and regulatory dependency across the Global South and proposes multipolar South-South AI governance involving the African Union and ASEAN.

Abstract

This paper examines structural power asymmetries in artificial intelligence development across the Global South, mapping dependencies across foundational compute, proprietary model APIs, data extraction, and regulatory capture. We propose multipolar policy frameworks connecting regional blocs such as the African Union and ASEAN through South-South cooperation to safeguard strategic autonomy and equitable technological futures.

Keywords: AI sovereigntyGlobal SouthAI governancegeopoliticsAAAI 2026South-South cooperation

Key Findings & Contributions

AAAI 2026
EGSAI Workshop
Philadelphia, USA
4 Vectors
Dependency Audit
Compute, APIs, Data, Policy
AU + ASEAN
Multipolar Alliances
South-South AI governance
Shared Pool
Strategic Agenda
Regional compute & datasets
  1. Categorizes structural Global South AI dependency into four reinforcing asymmetries: foundational compute monopoly, proprietary model API lock-in, raw data extraction, and external regulatory capture.
  2. Critiques single-bloc technology dependency and outlines multipolar South-South AI governance bridging the African Union (AU) and ASEAN.
  3. Proposes regional public-interest compute pooling, shared multilingual benchmarks, and reciprocity clauses for cross-border dataset licensing.

Benchmark Results

Four-Vector Structural Dependency Taxonomy & Multipolar Mitigation Levers

Dependency VectorCurrent Global South BottleneckRecommended Sovereignty Lever
Foundational ComputeZero domestic hyperscale clustersRegional federated academic HPC pools
Model & API LayerForeign closed-weight API lock-inOpen-weight regional domain adaptation
Data FlowsUncompensated raw linguistic extractionCommunity data trusts & benefit-sharing
Regulatory DesignVerbatim copy of external AI actsContext-tuned AU/EAC-ASEAN harmonization

BibTeX

@inproceedings{davies2026sovereignty,
  title={AI Sovereignty in the Global South: Power, Dependency, and Strategic Futures},
  author={Davies, S. and Nagaraja, V. H. and Nyalala, Innocent and Bhatt, N. and Banerjee, S.},
  booktitle={AAAI 2026 Workshop on Equitable and Sustainable Global AI (EGSAI)},
  address={Philadelphia, USA},
  year={2026},
  month={January}
}
APA Citation

Davies, S., Nagaraja, V. H., Nyalala, I., Bhatt, N., & Banerjee, S. (2026, January). AI Sovereignty in the Global South: Power, Dependency, and Strategic Futures. AAAI 2026 Workshop on Equitable and Sustainable Global AI (EGSAI), Philadelphia.