Data Is Nothing Like Oil: Dusting Off Copyright’s Originality Doctrine for the AI Age

Key findings from this study

• The authors propose that data should be conceptualized as a public good rather than a finite resource analogous to oil.
• The study identifies that copyright law's originality doctrine provides a more appropriate analytical framework for regulating data in AI contexts than commodity-based models.
• The authors demonstrate that current copyright disputes over AI training data reveal significant gaps between existing legal doctrine and technological practice.

Overview

Copyright law's originality doctrine requires examination for AI-driven data use contexts. The paper challenges the analogy equating data with oil, emphasizing that data constitutes a public good characterized by non-rivalrous and non-excludable properties. Current copyright frameworks face testing through litigation concerning training data derived from copyrighted works without owner consent. The authors argue that copyright doctrine should reconceptualize data more as an ocean resource than an extractable commodity.

Methods and approach

The analysis employs conceptual and legal framing to distinguish data from oil through economic and legal characteristics. The authors examine copyright law's originality doctrine as a lens for understanding data's status. They review the landscape of emerging copyright litigation globally, government consultation responses, and private licensing negotiations between AI firms and content industries. The framework considers how metaphorical understanding shapes legal and policy responses to AI training data practices.

Results

The originality doctrine provides a more coherent framework for data than commodity extraction models. Data differs fundamentally from oil in tangibility, finiteness, and economic excludability. Copyright infringement disputes surrounding AI training datasets reveal tensions between current law's application and technological practices. Private licensing arrangements between large technology firms and content industries reflect attempts to negotiate exclusive access outside formal legal structures. The metaphorical framing of data shapes institutional responses, including government policy development and litigation strategies.

Implications

Recalibrating copyright doctrine around originality may resolve conflicts between AI development and copyright protection. The public good nature of data suggests regulatory approaches should differ from those governing finite resources. Legal frameworks must account for data's fundamentally different characteristics from tangible commodities to ensure coherent policy outcomes. The ongoing negotiations and litigation indicate that metaphorical understanding directly influences practical legal and commercial arrangements.

Scope and limitations

This summary is based on the study abstract and available metadata. It does not include a full analysis of the complete paper, supplementary materials, or underlying datasets unless explicitly stated. Findings should be interpreted in the context of the original publication.