The Algorithmic Rents Research Program

In the AI Story

The program emerged from Mazzucato's broader analytical framework distinguishing value creation from value extraction — a distinction her earlier work had applied primarily to finance and pharmaceuticals. Applying it to AI platforms required new conceptual tools because the mechanisms of extraction in the platform economy differ structurally from those in traditional industries. Platforms do not extract through monopolistic pricing alone. They extract through algorithmic control over the attention, data, and lock-in of users who depend on them.

The empirical work centered on documenting how platforms' algorithmic systems behave as their market positions consolidate. Early-stage platforms compete for users by creating genuine value — better matching, cheaper prices, improved user experience. Mature platforms, having achieved network effects and switching-cost lock-in, deploy the same algorithmic systems to capture value from captive users. The same technology that produced value creation in the competitive phase produces algorithmic rents in the monopolistic phase. The transition is gradual, invisible to users, and structurally aligned with the platform's fiduciary obligations to shareholders.

The program's application to AI specifically illuminates a dynamic the democratization narrative obscures. Users pay subscription fees for AI capability while also providing data, attention, and lock-in that accumulate into the platform's competitive advantage. The financial terms appear favorable — twenty to two hundred dollars per month for extraordinary productivity gains — but the full cost of the transaction includes dimensions that do not appear on the invoice.

At the 2023 Algorithmic Rents Research Showcase, Mazzucato drew the explicit connection between rent extraction and the migration of AI researchers from public institutions to private companies. The compensation differentials that drive the brain drain are funded, in significant part, by revenues derived from algorithmic control over users rather than from genuine technological innovation.

Origin

The research program built on Mazzucato's decade-long collaboration with economists and legal scholars studying the structural features of digital platforms. Tim O'Reilly, the technology publisher and long-time observer of Silicon Valley, contributed the industry expertise and the phrase creating more value than you capture that became a touchstone of the program's normative framework. Ilan Strauss, a development economist trained in heterodox traditions, contributed the empirical methodology for measuring extraction against creation.

The program's publications in Data & Policy (2024) and the Hastings Science & Technology Law Journal established the academic foundation for policy proposals that Mazzucato subsequently advanced in her commentary on AI governance — including the proposals for windfall taxation, conditionality, and public AI infrastructure that animate the later chapters of her AI thinking.

Key Ideas

Algorithmic attention rents. The mechanism by which platforms extract value from users through algorithmic control over attention as market position consolidates.

Profits vs. rents distinction. Revenues from genuine innovation are profits; revenues from market power are rents — a distinction that tax policy and competition law should enforce.

Brain drain funding loop. The compensation drawing researchers from public institutions to private AI companies derives substantially from rent extraction, not from value creation.

Dual-phase platform behavior. The same algorithmic systems that produce value in competitive phases produce extraction in monopolistic phases — technology is not neutral with respect to market structure.

Policy implication. Competition law and taxation must distinguish between returns on innovation (to be encouraged) and returns on market power (to be constrained).

Debates & Critiques

The program has been criticized by mainstream economists for conflating different sources of platform profitability. Defenders of the platforms argue that what Mazzucato terms rents are in fact returns on the substantial R&D investments platforms make. The program's response is empirical: the profit margins of mature platforms exceed by large margins the returns that would be consistent with competitive R&D-driven innovation, suggesting that market power rather than innovation explains the gap.