Digital Empires: The Global Battle to Regulate Technology is a 569-page book by Anu Bradford, a professor at Columbia Law School and the originator of the "Brussels Effect" concept, published in September 2023 by Oxford University Press. It argues that the global digital economy is governed by three competing regulatory models — an American market-driven model, a European rights-driven model, and a Chinese state-driven model — each reflecting the geopolitics and political economy of its originator, and it maps how these models clash, coexist, and co-opt one another.
Three regulatory models
Bradford identifies three distinct regulatory approaches to the digital economy.
The American market-driven model has techno-libertarian roots and favors light-touch federal regulation. Bradford treats Section 230 as a structural feature that enabled platform growth, and describes state-level experimentation (California, New York) as an emerging second-best path. Enforcement runs primarily through antitrust litigation and state-attorney-general action rather than ex-ante rules. The model's associated exports are American tech platforms such as Google, Meta, Apple, and Microsoft.
The European rights-driven model is grounded in fundamental rights, expressed through instruments including GDPR, the DSA, the DMA, and the AI Act. It relies on ex-ante regulation targeting specific harms (data protection, algorithmic transparency, competition) and is enforced via DG CONNECT, national data protection authorities, and competition authorities. Bradford presents it as an explicit "third way" between the US and Chinese approaches. Its export mechanism is the Brussels Effect — the market-size-driven export of regulatory standards, the subject of Bradford's earlier work from 2012 onward — whereby EU rules such as GDPR and the AI Act impose compliance costs on the global operations of non-EU companies and thereby shape the non-EU internet.
The Chinese state-driven model is state-centric, with the CAC, MIIT, and party organs as primary regulators, and treats regulation as an instrument of state control through content moderation, surveillance, and social credit. Domestic tech champions (Alibaba, Tencent, Baidu, ByteDance) are governed in-house. Bradford characterizes Chinese regulation as marked by "verticalization" rather than "horizontalization" — sector-specific and often retroactive rules. The model is exported via the Digital Silk Road, which links BRI infrastructure and surveillance products to the Global South.
Interactions between the models
Bradford maps where the three models clash, coexist, and co-opt one another. The Brussels Effect extends EU rules such as GDPR and the AI Act to US companies by imposing costs on their global operations, so that EU regulation shapes the non-EU internet. Decoupling pressures — US export controls on China, China's self-reliance drive, and EU strategic autonomy — push toward a fragmented "splinternet." Third countries including India, Brazil, and ASEAN members face pressure to align with one model and often cherry-pick elements of several.
Relevance to AI regulation
Bradford's three-model framing serves as a reference lens for comparing EU, US, and Chinese approaches to AI regulation. The European rights-driven model maps onto instruments such as EU AI Act (Regulation 2024/1689), EU AI Office — Enforcement Framework + GPAI Provider Guidelines, EU General-Purpose AI Code of Practice (Final Version, 2025), UK AI Bill — Status and Delay (Source Summary), and Data (Use and Access) Act 2025 — Source Summary, with the EU AI Act (Regulation 2024/1689) serving as the paradigmatic rights-driven export. The American market-driven model maps onto America's AI Action Plan, Executive Order 14179 — Removing Barriers to American Leadership in Artificial Intelligence, Executive Order 14319 — Preventing Woke AI in the Federal Government, Executive Order 14320 — Promoting the Export of the American AI Technology Stack, California SB 53 — Transparency in Frontier AI Act, New York RAISE Act (S. 8828), Texas Responsible AI Governance Act (TRAIGA / HB 149) — Source Summary, Colorado AI Act (SB 24-205) and SB 25B-004 (Date Amendment), Illinois SB 3444 — Artificial Intelligence Safety Act, and State AG AI Guidances (CA, NJ, MA, OR). The Chinese state-driven model maps onto China — Interim Measures for the Management of Generative AI Services, China — Internet Information Service Algorithmic Recommendation Management Provisions, China — Provisions on the Administration of Deep Synthesis Internet Information Services, and China's Military AI Wish List. The cross-model conflicts Bradford describes are reflected in BIS Framework for AI Diffusion — Interim Final Rule (Jan 13, 2025), CSIS — DeepSeek, Huawei, Export Controls, and the Future of the U.S.-China AI Race (Allen, March 2025), Letter to Sec. Lutnick on H20 Restrictions (2025-07-28), and China and the US Are Running Different AI Races.
Confidence notes
The three-model framing carries high confidence as a comparative lens for EU-US-China tech regulation. Specific predictions about how the models interact carry medium confidence: the 2023 publication date precedes many 2024–2026 developments, including AI Act enforcement, the spread of US state AI laws, and refinements in Chinese AI industrial policy.
Relationships
- supports: Techno-Federalism: How Regulatory Fragmentation Shapes the U.S.-China AI Race — Wu's more recent framing specializes Bradford for AI specifically.
- supports: China and the US Are Running Different AI Races — empirical complement on China vs. US strategy.
- supports: Geopolitics in the Age of Artificial Intelligence — Sullivan & Feldman build on Bradford's framing.
- related: OpenAI — Industrial Policy for the Intelligence Age / The Tech High Ground (Jake Sullivan, Foreign Affairs) — US policy-competition arguments.
- related: EU AI Act (Regulation 2024/1689) — the paradigmatic "rights-driven empire" export.
- related: Incentives or Obligations? The U.S. Regulatory Approach to Voluntary AI Governance Standards — US voluntary-standards approach vs. EU mandatory.