Author(s)

Rodrigo Brandão, Glauco Arbix

Department of Sociology, Faculty of Philosophy, Letters and Human Sciences, University of São Paulo, São Paulo, Brazil.

The literature on using Artificial Intelligence (AI) systems for elaborating and implementing public policies is under development. Therefore, little is known about how the public sector can and should use AI responsibly and in what situations it has been able to do so. To partially fill these gaps, we investigated the use of facial recognition (FR) systems to combat fraud in discounts and gratuities guaranteed by law to specific groups, such as students and the elderly, in the 30 largest Brazilian municipalities. Based on primary data gathered through the Access to Information Law, we found that 14 use the technology in question for the investigated purpose, and we prepared the Facial Recognition Responsible Use Index (FRRU-I) and the Facial Recognition Responsible Use Scale (FRRU-S). None of the cities studied reached the “Very high” level in the FRRU-S; five of them achieved “High” level scores; three were at the “Intermediary” level; two obtained a “Low” score, and the remaining four showed a “Very low” score. These results suggest that the largest cities in the country do not have the necessary administrative structures for the responsible use of FR systems or do not know how to mobilise the structures they have to promote such use. Our findings also indicate that the public sector’s unfamiliarity with FR systems may extend to other AI applications.

Artificial Intelligence, Facial Recognition, Ethics, Public Policy, Public Transport

Brandão, R. and Arbix, G. (2022) Artificial Intelligence, Ethics and Public Policy— The Use of Facial Recognition Systems in Public Transport in the Largest Brazilian Cities. Journal of Service Science and Management, 15, 551-575. doi: 10.4236/jssm.2022.155032.