A Technical Typology of AI Systems in Public Administration

2026-06-30 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

A new technical typology categorizes artificial intelligence (AI) systems into five distinct types: hand-coded, glass-box, black-box, general-purpose, and agentic systems. This framework, specifically calibrated for public administration, aims to address the current lack of technical precision in research, which often treats "AI" as a monolithic entity. An analysis of 91 highly-cited public administration papers published between 2019 and 2025 revealed significant imprecision: 55% of studies underspecified the AI system, 31% motivated their work with a different system than they actually studied, and 41% drew conclusions broader than their studied system supported. The research provides practical recommendations for future studies, including a diagnostic guide to help researchers accurately classify AI systems based on publicly available information.

Key takeaway

For research scientists and AI ethicists studying public sector AI, you must adopt a more technically precise approach to classifying AI systems. Your current research may be drawing over-generalized conclusions or misrepresenting system impacts on public values like accountability. Utilize the proposed five-category typology and diagnostic guide to accurately specify the AI systems you analyze, ensuring your findings are robust and directly applicable to policy development.

Key insights

Technical precision in classifying AI systems is critical for public administration research to accurately assess impacts on public values.

Principles

• AI systems require technical distinctions for public value analysis.
• Research must avoid underspecifying AI system types.
• Conclusions should not over-generalize beyond the studied system.

Method

A five-category typology (hand-coded, glass-box, black-box, general-purpose, agentic) is proposed, calibrated for public administration, alongside a diagnostic guide for system classification.

In practice

• Classify AI systems using the five-category typology.
• Utilize the diagnostic guide for system identification.
• Ensure research scope matches the AI system studied.

Topics

• AI Systems Typology
• Public Administration AI
• Research Methodology
• Public Values
• Accountability
• AI Ethics

Best for: AI Scientist, Research Scientist, AI Ethicist, Policy Maker

Related on AIssential

• Societal Impacts — AI research must integrate human values, anticipate misuse, and inform policy to ensure beneficial societal outcomes.
• Bureaucratic Silences: What the Canadian AI Register Reveals, Omits, and Obscures — AI registers, like Canada's, act as ontological design instruments, shaping accountability by emphasizing technical aspects while obscuring crucial human and contextual factors.
• Charting the AI Perception Gap: Across 71 scenarios, AI experts (N=119) and the public (N=1100) have differing views on the risks, benefits, and value of AI. More importantly, AI experts discount the influence of risks stronger than the public does when forming their value judgments [R] — AI experts and the public hold systematically different mental models regarding AI's risks, benefits, and overall value.
• Digital monitoring is growing in South Africa’s public service – regulation needs to catch up — Rapid digital tool adoption in public sector evaluation outpaces ethical governance, creating risks for citizens and data integrity.
• Governments Are Using AI To Draft Legislation. What Could Possibly Go Wrong? — AI is being adopted by governments worldwide to enhance legislative efficiency, but raises significant concerns about legitimacy and trust.
• Book Review: Intellectuals and Society by Thomas Sowell — Intellectuals insulated from consequences can perpetuate bad ideas due to a broken feedback loop.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.