paperJune 2025UnreviewedOpen access

Oversight Structures for Agentic AI in Public-Sector Organizations

Chris Schmitz, Jonathan Rystrøm, Jan Batzner

Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025)

Abstract

This paper finds that the introduction of agentic AI systems intensifies existing challenges to traditional public sector oversight mechanisms -- which rely on siloed compliance units and episodic approvals rather than continuous, integrated supervision. We identify five governance dimensions essential for responsible agent deployment: cross-departmental implementation, comprehensive evaluation, enhanced security protocols, operational visibility, and systematic auditing. We evaluate the capacit

Cite

@inproceedings{schmitz2025oversight,
  title = {{Oversight Structures for Agentic AI in Public-Sector Organizations}},
  author = {Chris Schmitz and Jonathan Rystrøm and Jan Batzner},
  year = {2025},
  month = jun,
  booktitle = {Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025)},
  eprint = {2506.04836},
  archivePrefix = {arXiv},
  doi = {10.18653/v1/2025.realm-1.21},
  url = {https://www.semanticscholar.org/paper/f28ed25afc2d23d9d62e0766b2956d142bb9913a}
}