Reclaiming Feudalism for the Technological Era

Shelly Kreiczer-Levy

DOI: https://doi.org/10.2139/ssrn.4282789

2021-01-01

SSRN Electronic Journal

Abstract:Personal property law has a blind spot when it comes to technological items, as they do not account for the long-term, unequal property collaboration that is required in operating these assets. I argue that we can learn from the intellectual legal history of feudalism about the vulnerabilities produced by property collaborations between unequal parties. Owners of robots as AI objects (e.g., autonomous vehicles, drones, and robot-chefs) have limited control over their property. Users own the physical product, but they only have a license to use the software. As per the terms of the license, the manufacturer retains control over many aspects of the object’s ongoing use. Although this structure is criticized in the literature, none of the critics points out the need to rethink the current structure of these rights. AI products have autonomous decision-making capabilities that make their actions hard to foresee and require periodical updates to secure their safety and quality. This Article is the first to offer a property model for technological property collaborations. The inspiration for this property model lies in the historical form of feudal property. Feudalism is often evoked in the law and technology literature to warn us against the power that large corporations hold over users. While these concerns are valuable, I maintain that feudal property has important potential to identify and address the unique vulnerabilities in these property collaborations. First, the duties of users and manufacturers in robots as well as feudal property are not connected to the use and function of the asset. Second, the property can only be used with the collaboration of the manufacturer or lord. Following this analysis, this Article offers two models for property collaborations in AI products: a moderate, connection model and a more radical, competition model. The connection model adopts the basic feudal concept of split ownership accompanied by a specifically tailored relational role and applies it to robots with the necessary changes. The competition model seeks to create a market where different manufacturers compete for the development value of the robot. The proposed models have several normative implications, including invalidating limitations on use, justifying a right to repair and data portability, and clarifying the copyright protection of AI-produced creative work.