Link to paperAbstract:
Digital platforms, which control unique access points to the rich data stored on their servers, have become a “living lab” of real-time information. Scientists and researchers increasingly use platform data for various purposes, such as training machine learning (ML) systems and Natural Language Processing (NLP) models, and for studying diverse fields such as medicine, humanities, and social sciences, including the influence of digital platforms on society. However, researchers increasingly encounter significant barriers, when attempting to access platform data. Although platforms typically lack proprietary rights over the data itself, they exert strong control over its use by imposing digital locks and boilerplate contractual limitations. Faced with the legal risk of potential breach-of-contract lawsuits filed by well-funded platforms, researchers may simply opt to steer clear of platform data research.
This Article proposes private-law-centered solutions to overcome platform data lockout. First, researchers who access and use platform data without explicit permission should be able to contest breach-of-contract claims made against them by claiming copyright preemption. Platform data falls under copyright law, either because it is protected by copyright (such as user-generated-content) or because it constitutes basic "building blocks", such as users’ digital data trails, which are specifically excluded from copyright protection. When platforms robustly ban any reproduction of data, they effectively benefit from quasi-copyright protection, through private ordering, albeit compromising fundamental copyright principles, including fair use. Their contractual claims should, therefore, be preempted by copyright law. Second, courts should facilitate platform data research by narrowly interpreting boilerplate contractual bans on data access. Third, nuisance law may further support platform data research by empowering researchers to demand the removal of technological barriers that hinder access to public, non-proprietary data.
Private law solutions to platform data lockout, however, do not grant researchers an affirmative right to use platform data for research. Legislative action of the type recently pursued by the European Union is required to establish such a Right to Research. This Article therefore concludes by examining regulatory approaches to platform data lockout, concluding that combining private law solutions with regulatory intervention offers the most effective means of adequately facilitating platform data research.
Open Zoom
