This paper presents a qualitative audit of AncestryDNA's Ethnicity Estimate system using the Algorithmic Ecology framework. It examines how direct-to-consumer genetic testing services impact racialized communities and perpetuate historical patterns of datafication and racial classification.
The audit employs the Algorithmic Ecology framework developed by the Stop LAPD Spying Coalition and Free Radicals (2020), examining the algorithm's impact across four levels:
- Community
- Operational
- Institutional
- Ideological
See here for more information: Stop LAPD Spying Coalition's Algorithmic Ecology Framework
If you use this work, please cite: Clark (2024). "Auditing Ancestry: Considering the Algorithmic Ecology of the AncestryDNA Ethnicity Estimate." Oxford Internet Institute, University of Oxford. Fairness, Accountability and Transparency in Machine Learning, Hilary Term 2024.