From Regulatory Vacuum to Socially Shaped Governance: A Responsibility-Gap Model of Facial Recognition

Abstract

Facial recognition has transformed unchanging physical characteristics into instruments of economic and governmental power, but regulation has not kept pace with deployment; thus, enterprises are resorting to superficial ethical behavior. We believe that strategic underinvestment is an inevitable consequence of structural regulatory lag, not a moral failure by individuals, and dispersed social actors can help to address this problem. A delayed stock-flow system is employed to construct an ethical signalling game and a polycentric detection framework with correlated nodes for the construction of a formal responsibility-gap model. We apply theory-guided process tracing to the first facial recognition lawsuit in China (2019-2021) and, in conjunction with a five-regime fuzzy-set comparison, use a reproducible open dataset. We have built a socially constructed path of governance, and litigation, media dissemination, judicial decisions and legislative codification are all non-Western regulatory systems. The first is to extend the scope of ethics-washing to include the omission of under-investment, and the second is to differentiate between an institutional responsibility gap and a traditional epistemic gap.