Paper Type

Complete

Abstract

As governments worldwide race to govern artificial intelligence (AI), no validated instrument yet exists for systematically comparing their approaches. We address this gap by developing a six-dimensional text-based governance index and applying it to 72 policy documents across eight leading AI jurisdictions from 2016 to 2025. The index passes all temporal validity checks and aligns with established benchmarks, confirming it captures meaningful governance variation. We provide evidence that specific governance configurations are consistently associated with strong AI ecosystem outcomes. In particular, innovation framing combined with regulatory substance is consistently sufficient for high private AI investment across diverse institutional traditions. Importantly, no single governance dimension is individually necessary or sufficient. By extending AI governance measurement from organizational to national-level, we demonstrate that the coherence of the governance mix matters more than any individual policy lever. These findings offer timely guidance for IS researchers and policymakers shaping the future of AI governance.

Paper Number

1480

Comments

SIG E-GOV

Download

Share

COinS
 
Aug 15th, 12:00 AM

Rethinking AI Governance: Policy Configurations and Ecosystem Performance

As governments worldwide race to govern artificial intelligence (AI), no validated instrument yet exists for systematically comparing their approaches. We address this gap by developing a six-dimensional text-based governance index and applying it to 72 policy documents across eight leading AI jurisdictions from 2016 to 2025. The index passes all temporal validity checks and aligns with established benchmarks, confirming it captures meaningful governance variation. We provide evidence that specific governance configurations are consistently associated with strong AI ecosystem outcomes. In particular, innovation framing combined with regulatory substance is consistently sufficient for high private AI investment across diverse institutional traditions. Importantly, no single governance dimension is individually necessary or sufficient. By extending AI governance measurement from organizational to national-level, we demonstrate that the coherence of the governance mix matters more than any individual policy lever. These findings offer timely guidance for IS researchers and policymakers shaping the future of AI governance.