Frank Fagan (South Texas College Law Houston) has posted “Training Data Governance” (NYU Journal of Intellectual Property & Entertainment Law, Vol. 15, pp. 227-64 (2026)) on SSRN. Here is the abstract:

As AI-generated summaries increasingly displace traditional search results, users are less likely to visit the underlying websites where content is published. This shift has sharply reduced traffic to those sites, threatening the economic viability of content creators and prompting a wave of paywalls, restrictions, and litigation. With referral-based revenue in decline, the continued supply of high-quality content faces mounting risk, precisely as generative AI has grown more dependent on such material. This tension, between innovation and sustainability, frames the central legal and policy inquiry of training data governance: how to preserve access to essential AI training inputs without undermining the incentives to produce them.

This Article examines licensing as a tool of training-data governance and focuses on the practical question of when content loss threatens model performance and reduces social welfare. The inquiry centers on whether withdrawal of high-value material is likely and whether voluntary bargaining can realistically prevent it. Where the risk of withdraw is low, additional protection is unnecessary; where the risk is substantial and bargaining fails, a narrowly tailored fallback, such as a standardized, non-exclusive license, can preserve access without disturbing fair-use doctrine or existing private arrangements. This welfare logic is implemented through a three-part test: licensing is warranted only when (1) the content has demonstrable value for AI training, (2) withdrawal is the rational market outcome absent remuneration, and (3) voluntary licensing fails due to transaction costs or bargaining frictions. Together, these conditions ensure that intervention occurs only where it improves overall welfare relative to the status quo.

Chiara Gallese (Tilburg U Tilburg Institute Law) and Elena Falletti (Carlo Cattaneo LIUC U) have posted “Dataset Assessment Tools for the AI Act Compliance of High-Risk AI Systems” (https://doi.org/10.13135/2785-7867/12840) on SSRN. Here is the abstract:

The European Union’s Artificial Intelligence Act (AI Act), effective as of August 2024, introduces stringent data governance requirements for high-risk AI systems. Central to these obligations is the mandate for data used in training, validation, and testing to be subject to rigorous quality controls and governance practices aligned with the system’s intended purpose. This paper examines the AI Act’s novel emphasis on responsible data handling as a foundation for mitigating bias, ensuring safety, and protecting fundamental rights. It argues that while existing scholarship largely focuses on algorithmic fairness at the model level, insufficient attention has been given to fairness embedded within the data itself. Addressing this gap, the paper explores the complex ethical and epistemological tensions inherent in defining “fairness”, and critiques current policy frameworks for their narrow focus on technical fixes. The study highlights the FAnFAIR tool as a pioneering approach to data fairness evaluation, offering a hybrid methodology that integrates automated statistical diagnostics with qualitative, context-aware assessments. FAnFAIR aligns its features with the AI Act’s legal mandates and facilitates compliance through continuous bias monitoring, data quality enhancement, and informed decision-making about dataset suitability. This paper provides a socio-legal and technical analysis of data governance under the AI Act, advocating for a shift toward embedded ethical oversight throughout the AI lifecycle and, in particular, in the data processing phase.

Neil M. Richards (Washington U Law) has posted “Privacy Is Not Theft” on SSRN. Here is the abstract:

Something curious has happened to privacy recently. In our post-pandemic age of generative AI, tech companies and governments have stopped asking for our data or relying on fictions of implied consent. They have simply started taking it, arguing that data collection is inevitable, necessary, and virtuous. As masked federal agents prowl American streets sticking facial recognition scanners in the faces of suspected undocumented aliens and as AI companies lament the lack of personal data to train their foundation models, there is a sense that some think privacy is getting in the way of the future. In their view, protections for human data are a kind of theft against convenience, a theft against safety, and a theft from the stockpile of personal data our machines need to create a frictionless future of leisure, abundance, and entertainment.

We’ve seen this idea before, but until recently only in the realms of farce and dystopian satire. Dave Eggers’ 2013 novel The Circle envisioned a sinister social network—a kind of Facebook on steroids—whose founders espoused a series of silly and dystopian mantras, including the idea that “Privacy Is Theft.” In the novel, the company uses these principles as the basis for their business, and by extension, a radically anti-privacy reworking of near-future society. Foreseeably dystopian hijinks ensue. Life imitates art, indeed, as a series of products and practices in our non-fictional world have begun to reflect the mantra that “Privacy Is Theft,” suggesting that indiscriminate data collection is both (1) inevitable and (2) virtuous.

This essay uses Jonathon Penney’s new book on chilling effects in connection with the broader trans-disciplinary and trans-Atlantic scholarship on what privacy is and why it matters to critique the way some powerful entities have started to talk about privacy as a kind of theft.  Penney’s work shows the hollowness of these curious suggestions that privacy is no longer a fundamental right but rather a kind of theft from the future, or at least a theft from people who think they are building a most excellent future for themselves and want your information as raw material to do that. The essay presents a few recent technological developments, explains the ideology of Privacy-as-theft that underlies them, and suggests that Penney’s work allows us to better deconstruct that ideology and helps us to understand why they it is problematic, and how the law is failing us—at least if we still care about things like democracy, eccentric individuality, and the development of new, heretical ideas. My claim here is a simple one—that privacy is not actually theft, but an essential value that we need to protect if we want to continue to live in societies that preserve individual rights, self-government diversity, and just plain eccentric weirdness. The real thieves are the ones who want your data and complicity and who are building unnecessary tools to steal it.

My argument proceeds in four steps. Part I presents the ideology of Privacy As Theft in The Circle. Part II discusses three new technological developments embodying this shift—Meta’s aggressive marketing of its “smart” glasses, the use of facial recognition tools by ICE personnel seeking undocumented immigrants, and Zoom’s presumption that all meetings should be transcribed and assessed by its artificial intelligence tools. Part III explores Penney’s work on chilling effects in the context of the broader literature, showing how at the level of theory and at that of empirical social science, privacy is necessary to any society that aspires to individuality, eccentricity, and/or political freedom. Part IV shows how the ideology of Privacy As Theft is not just hogwash, but hogwash in pursuit of a much greater theft of its own—the power to dictate what kind of society we will inhabit in the future.