Algorithmic Administration as Constitutional Governance

ORCID

https://orcid.org/0000-0002-5496-2104>Cary Coglianese 0000-0002-5496-2104

Document Type

Article

Publication Date

9-19-2024

Abstract

Government agencies are increasingly adopting digital technologies that rely on AI to aid their work. The algorithmic automation of administrative functions and decision-making—including in law enforcement, adjudication, delivery of public services, legal guidance, and inspections—offers a range of benefits in terms of improving the performance of government. But the use of AI tools by government also raises a host of concerns about bias, transparency, and accountability. So far, most academic research in law has focused on these concerns about the use of AI, with some scholars urging resistance to governmental use of AI by raising constitutional objections. This chapter suggests an alternative connection between constitutionalism and governmental use of AI tools. Rather than seeing the public as having constitutional rights to be free from AI in government—such as by having a so-called right to a human-made decision—the authors suggest that the public may well have certain claims sounding in constitutional values to have government institutions make use of AI tools. Accepting this alternative account need not require the dismissal of all concerns about how governments could use AI to infringe on human rights or create serious problems. Rather, it suggests only that government authorities need to deploy AI with care. The chapter concludes by outlining principles and practices for the responsible use of AI in the administrative state.

Keywords

public administration, government agencies, automation, artificial intelligence, effective government, digital constitutionalism, constitutional rights

Publication Title

Oxford Handbook on Digital Constitutionalism

DOI

https://doi.org/10.1093/oxfordhb/9780198877820.013.27

Link to Full Text

Share

COinS