Corporations are increasingly using technology to conduct business, seeking greater automation and efficiencies while decreasing costs. Indeed, several states are considering changes to their business-formation laws to accommodate completely automated businesses – those run through, or by, self-executing computer code and artificial intelligence. Internationally, several jurisdictions already offer corporation-equivalent business structures to completely automated businesses. Together, these developments set the stage for a world where autonomous business organizations enjoy the same rights and responsibilities as corporations – a world in which an autonomous organization enjoys the legal fiction of personhood. In a new article, Autonomous Corporate Personhood, I use autonomous corporations as a case study to argue that building a comprehensive legal approach to artificial rights – rights enjoyed by artificial “people,” whether entities, machines, or both – requires consideration of the varied socio-technical contexts in which artificial people exist.
Because autonomous corporations are composed of both a corporate entity and an artificially intelligent system (“AI System”), considering the nature of autonomous corporate personhood starts with the discussion among academics and policy makers as to whether and to what extent to treat AI Systems as persons. The issue often arises in the context of creating a liability and accountability structure for AI Systems or in discussions about how to deter the potential negative societal impacts of anthropomorphiizing AI Systems. A review of the AI Systems literature suggests a spectrum of personhood for AI Systems that varies by levels of automation and by context.
On the one hand, an AI System that automates activity without exercising any form of will might appropriately be viewed as property of another actor that uses the AI System as a mere tool. On the other hand, highly autonomous AI systems might represent a form of hybrid social person – one to which natural persons attach a level of social meaning but that does not possess characteristics normally associated with natural persons. And in between these two extremes, AI Systems and other emerging technologies act as a conduit for collective human action. This middle group remains underexplored, with important questions persisting about whether the AI System substitute carries forward the personhood attributes of the human that acts through it.
This is the point at which autonomous corporations offer a vehicle for teaching deeper lessons about legal personhood. Both corporations and AI Systems are artifacts in the sense that they are both technologies to which the law can, and sometimes does, attach certain legal fictions. Artifacts, for their part, do not exist in a vacuum but sit and act within a social context. Considering the use of AI Systems within the social context of the corporation offers a unique opportunity to explore the use of artifacts by humans as a vehicle for collective action.
A review of the current uses of AI Systems in business uncovers a striking variety of approaches. Some corporations use AI Systems and other automating technologies as mere tools in running a business (Traditional Plus). Other businesses use AI systems and other emerging technologies as a device to coordinate operations and provide incentives to workers (Distributed Business Entities). When used this way, the AI System clearly takes action that moves beyond mere property, such as autonomous governance or autonomous compliance, but it does not exhibit emergent characteristics or act on its environment of its own accord. The most fully autonomous organizations (Autonomous Entities) currently in existence interact with natural persons, posing some of the same policy concerns as socially valent AI Systems.
Because an autonomous corporation is both AI System and corporation, a look at personhood must also consider the three traditionally dominant theories of corporate personhood – artificial entity theory, the aggregate theory, and the real entity theory. Together with the AI System personhood approaches, the corporate personhood theories help construct an approach to the core question of autonomous corporate personhood. Namely, when should a corporation that uses AI Systems receive legal personhood, and what should the scope of that personhood encompass?
For Traditional Plus corporations, the answer emerges from the fact that, when a traditional corporation merely uses an AI System as a cost-savings tool, it does not (and should not) alter the corporation’s status under traditional corporate personhood doctrine. Currently, corporations enjoy only the benefits of personhood given to them by statute and the U.S. Supreme Court, not the full set of rights enjoyed by humans. As a result, we might think of this as “restricted personhood.” When a Distributed Business Entity, for its part, fulfills all of the legal formalities of a corporation and yet enables radically flatter governance structures, it closely tracks theories of the corporation as an aggregate of natural persons and of AI Systems as conduits for collective human activity. Under such circumstances, strong policy arguments exist to offer Distributed Business Entities fuller personhood rights than Traditional Plus corporations (full personhood). Autonomous Entities, however – corporations not controlled or perhaps not even owned by humans – almost literally embody conceptions of corporations as an artificial entity. Such Autonomous Entities will also act meaningfully in society in ways that may require protection of natural people – like socially valent robots. Thus, autonomous corporate personhood for Autonomous Entities might fit a “limited personhood” concept in which personhood is a tool for protecting humans from the autonomous corporation, and preventing humans from abusing the autonomous corporate form.
In other words, viewing autonomous corporations as a system composed of two artifacts – AI System and corporation – calls for applying both areas of personhood theory to autonomous corporate personhood. Doing so reveals a spectrum of personhood that varies by the social context in which the artifact sits. This autonomous corporate personhood spectrum demonstrates that when considering the nature and scope of legal personhood for artifacts – whether corporation or AI System – context should play a significant role in defining the appropriate bundles of rights and duties that attach. The key to understanding the context lies in properly assessing both the socio-technical reality of the AI System and the socio-legal reality of the use to which the AI-System is put. Importantly, tying legal regimes for artifacts to different socio-technical contexts enables the development of rules that focus on function rather than on specific technology.
Doing so also emphasizes that, when it comes to appropriate rules for legal personhood, one single theory of personhood may not be able to rule all scenarios, even within one specific context like autonomous corporations. Ultimately, the investigation into autonomous corporate personhood cautions that, at the intersection of law and technology, without deeper investigation into both the reality of the technology and the legal demands of the social context in which the technology will be used, legal rules risk continuing generalizations and perpetuating myths that exacerbate the extent to which law lags behind or, worse, compounds the risks related to technology.
 Wash. L. Rev. (forthcoming 2021).
 For a deeper dive into the spectrum of uses and how they might be grouped in analytically useful ways, see Carla L. Reyes, Autonomous Business Reality, 20 Nev. L. J. _ (forthcoming 2021).
 Emergence refers to the capacity of certain AI Systems to act in useful but unexpected, unplanned, and sometimes inexplicable ways. See Ryan Calo, Robotics and the Lessons of Cyberlaw, 103 Cal. L. Rev. 513 539-40 (2015).
This post comes to us from Professor Carla L. Reyes at Southern Methodist University’s Dedman School of Law. It is based on her recent paper, “Autonomous Corporate Personhood,” forthcoming in the Washington Law Review and available here.