Navigating Human Identity in AI Algorithmic Bias and Reasonable Technology Deployment in Environmental Impact Assessment

Human identity plays a dominant role in the discourse on algorithmic bias in artificial intelligence at this very moment. It is precisely the very attributes that constitute identity- race, caste, gender, disability, class position and locational belonging which are transmuted into data points, proxy variables and risk indices in algorithmic systems, and often without the consent or even a decent understanding of the affected persons. In the field of Environmental Impact Assessment (EIA), where AI tools are increasingly the basis for decision support tools or models of pollution dispersion, health risk prediction, mitigation prioritization and ranking of project alternatives, how identity is encoded or not encoded in data-sets thus affects who is made visible as “at risk” and who is effectively excluded from environmental decision making.