Government Also Uses AI — Is 'AI Government' Efficiency or Automated Control?

The government is introducing AI comprehensively across all administration. The Ministry of Science and ICT presented plans to use AI across the entire administrative lifecycle — report writing, press releases, public complaint response, budget review — on a common government-wide infrastructure. The government characterizes this as implementing a "well-working AI government." But when AI enters administration, the questions are not simple: is AI government a tool for improving efficiency, or a new power that automates control methods? The structural change: standardizing and automating common public administration tasks through AI; AI applied first to report draft writing, policy explanation material organization, and public consultation; civil servant role shifts from "writer" to "reviewer." The responsibility gap risk: AI-generated documents and analysis become de facto baselines; if civil servants'' judgments occur within AI''s framework, who bears responsibility when errors occur? The AI, the civil servant who used it, the government that designed the system, or the company that provided the technology? The "invisible control" dimension: AI government''s most significant change is not speed but the method of surveillance and evaluation; government plans to manage all administrative processes quantitatively through integrated data and AI analysis — creating a structure that simultaneously reduces administrative burden and enables real-time tracking of civil servants'' work patterns and performance; formal reporting may decrease while informal behavioral control increases. Transparency paradox: AI-generated policy analysis could be more objective (consistent application of criteria) or less transparent (algorithmic decisions harder to audit than human reasoning chains) depending on design choices that remain unspecified. The accountability framework requires: clear designation of human decision-makers who review and are responsible for AI-assisted decisions; audit trails of AI inputs and outputs; appeal mechanisms for decisions influenced by AI analysis; and explicit policy on when AI recommendations can be overridden.