Automated Disinformation and Malicious AI Swarms: Risks for Democracy and Development in Africa

This preprint examines how coordinated AI agents could support disinformation campaigns in African contexts, with a focus on Mali and Ethiopia. It considers how political institutions, media conditions, language coverage, and human involvement shape both potential threats and defensive capacity. The analysis distinguishes malicious AI swarms from simpler uses of AI to generate content or manage synthetic accounts.

The paper presents a forward-looking risk analysis, not evidence that fully autonomous influence swarms already operate at scale in Africa. It proposes safeguards combining technical measures, platform accountability, stronger civic institutions, and regional cooperation.

Read the paper on arXiv

Bibtex

@misc{schroeder2026automateddisinformation,
  title={Automated Disinformation and Malicious AI Swarms: Risks for Democracy and Development in Africa},
  author={Schroeder, Daniel Thilo and Lutscher, Philipp M. and Badji, Samba Dialimpa and Brenner, Stefan and Dong, Wenchao and Haidemariam, Tsehaye and Ibrahim, Saheed Bidemi and Kebede, Amanuel Tesfaye and Kunst, Jonas R. and Langguth, Johannes and Lemaire, Pauline and Moges, Mulatu Alemayehu and Olejnik, Lukasz and Orgeret, Kristin Skare and Walulya, Gerald},
  year={2026},
  eprint={2610.11930},
  archivePrefix={arXiv},
  primaryClass={cs.CY},
  doi={10.48550/arXiv.2610.11930}
}