Artificial intelligence, task complexity and uncertainty: analyzing the advantages and disadvantages of using algorithms in public service delivery under public administration theories

Abstract

PurposeThis article revisits some theories and concepts of public administration, including those related to public value, transaction costs and social equity, to analyze the advantages and disadvantages of using artificial intelligence (AI) algorithms in public service delivery. The author seeks to mobilize theory to guide AI-era public management practitioners and researchers.Design/methodology/approachThe author uses an existing task classification model to mobilize and juxtapose public management theories against artificial intelligence potential impacts in public service delivery. Theories of social equity and transaction costs as well as some concepts such as red tape, efficiency and economy are used to argue that the discipline of public administration provides a foundation to ensure algorithms are used in a way that improves service delivery.FindingsAfter presenting literature on the challenges and promises of using AI in public service, the study shows that while the adoption of algorithms in public service has benefits, some serious challenges still exist when looked at under the lenses of theory. Additionally, the author mobilizes the public administration concepts of agenda setting and coproduction and finds that designing AI-enabled public services should be centered on citizens who are not mere customers. As an implication for public management practice, this study shows that bringing citizens to the forefront of designing and implementing AI-delivered services is key to reducing the reproduction of social biases.Research limitations/implicationsAs a fast-growing subject, artificial intelligence research in public management is yet to empirically test some of the theories that the study presented.Practical implicationsThe paper vulgarizes some theories of public administration which practitioners can consider in the design and implementation of AI-enabled public services. Additionally, the study shows practitioners that bringing citizens to the forefront of designing and implementing AI-delivered services is key to reducing the reproduction of social biases.Social implicationsThe paper informs a broad audience who might not be familiar with public administration theories and how those theories can be taken into consideration when adopting AI systems in service delivery.Originality/valueThis research is original, as, to the best of the author’s knowledge, no prior work has combined these concepts in analyzing AI in the public sector.