Title: GovAI-Pipe: Establishing a Multi-Tiered AI Governance Framework for Citizen-Centric Services in Turkey’s e-Government Portal

Abstract

Turkey’s e-Devlet platform, a centralized e-government gateway, currently supports over 68 million registered users across more than 9,200 distinct government services. As the system increasingly incorporates artificial intelligence into citizen-facing interfaces—such as chatbot support and eligibility screening—there remains a critical gap in structured technical infrastructure. Specifically, no existing mechanism effectively bridges high-level policy mandates, including the EU AI Act, the OECD AI Principles, and Turkey’s National AI Strategy, with the practical deployment of AI within this centralized environment.

To address this disconnect, we introduce GovAI-Pipe, a four-tier governance pipeline developed through Design Science Research. This framework aligns the AI model lifecycle with specific governance checkpoints: (1) pre-deployment validation, which includes bias testing, explainability checks, and privacy impact assessments; (2) deployment governance, featuring risk-tier classification and approval workflows; (3) runtime monitoring, focused on drift detection, fairness tracking, and human-in-the-loop escalation protocols; and (4) post-incident governance, ensuring audit trails, system rollback capabilities, and citizen redress mechanisms. Each stage of the pipeline is mapped to relevant provisions within the EU AI Act, the GDPR data protection standards, and the National AI Strategy. We validate the framework’s efficacy by applying it to two high-risk e-Devlet use cases, demonstrating how GovAI-Pipe transforms abstract governance principles into auditable, technical pipeline components.

Source: arXiv Generated at: 2026-06-02 00:00:00 UTC