Individual Submission Summary
Share...

Direct link:

Operationalizing Corporate Quasi-Sovereignty in AI Governance: A Multi-Stream Typology and Methodological Framework

Saturday, November 7, 3:30 to 5:00pm, Property: Boston Marriott Copley Place, Floor: 3rd Floor, Room: Suffolk

Abstract

A central question in AI governance research is how private sector actors shape regulatory outcomes in ways that transcend conventional lobbying or market influence. While previous scholarship has documented the growing power of technology firms in shaping AI policy, the systematic conceptualization and measurement of how AI-driven firms exercise governance authority resembling sovereign state influence remains underdeveloped. As AI technologies advance and their societal implications intensify, understanding the structural and behavioral determinants of corporate quasi-sovereign power is crucial for anticipating governance outcomes and designing effective regulatory responses.

To address this gap, we draw on the concept of corporate quasi-sovereignty (CQS), which captures the capacity of AI firms to shape governance in ways that parallel state authority. While existing treatments of CQS remain largely theoretical, this paper operationalizes the framework through a multi-stream typology grounded in Kingdon's (1984) multiple streams theory. We focus on two streams as primary determinants of quasi-sovereign capacity. The policy stream captures firms' infrastructural and market power across three dimensions: technological assets (AI patents, R&D spending, proprietary infrastructure), data resources (volume, variety, and velocity of data holdings), and platform dominance (user base size, third-party dependencies, international reach). The political stream captures CEO power within the firm (founder status, voting rights, board control) and entrepreneurial orientation toward AI governance (political donations, policy engagement, thought leadership, and strategic coalition-building).
The intersection of these two streams produces a two-dimensional typology of four CQS ideal types. Infrastructural Titans combine high policy and political capacity, leveraging vast technological resources alongside politically influential leadership to shape both technical standards and policy debates. Data Dynamos possess significant infrastructural power but less politically active leadership, relying on market-based mechanisms to influence governance. Entrepreneurial Evangelists hold more limited technological resources but exercise disproportionate governance influence through CEO thought leadership and coalition-building. Niche Navigators operate with specialized capabilities and limited political profiles, concentrating influence within specific sectors or domains.

To demonstrate empirical utility, we apply the typology to U.S.-based AI firms listed on the Nasdaq stock exchange, drawing on patent databases, company filings, web traffic analytics, campaign finance disclosures, and executive profiles to classify firms across the four configurations and analyze their distributional characteristics.
The paper makes several contributions. Theoretically, it extends multiple streams theory to the corporate governance context, demonstrating how policy and political stream dynamics generate distinctive quasi-sovereign configurations at the firm level. Methodologically, it offers a replicable measurement framework specifying data sources, indicators, and thresholds for each construct. Empirically, it provides the first systematic cross-firm mapping of quasi-sovereign capacity among the most consequential AI firms in the U.S. context. Together, these contributions equip scholars and policymakers with actionable tools for understanding how private sector power shapes AI governance and identifying where regulatory interventions are most urgently needed.

Author