Three Laws Blog: Law and Artificial Intelligence

Craig and Kerr on the Death of the AI Author @CraigCarys @ianrkerr

Carys J. Craig, Osgoode Hall Law School, and Ian R. Kerr, University of Ottawa, Common Law Section, have published The Death of the AI Author. Here is the abstract.

Much of the second-generation literature on AI and authorship asks whether an increasing sophistication and independence of generative code should cause us to rethink embedded assumptions about the meaning of authorship, arguing that recognizing the authored nature of AI-generated works may require a less profound doctrinal leap than has historically been suggested. In this essay, we argue that the threshold for authorship does not depend on the evolution or state of the art in AI or robotics. Instead, we contend that the very notion of AI-authorship rests on a category mistake: it is not an error about the current or potential capacities, capabilities, intelligence or sophistication of machines; rather it is an error about the ontology of authorship. Building on the established critique of the romantic author figure, we argue that the death of the romantic author also and equally entails the death of the AI author. We provide a theoretical account of authorship that demonstrates why claims of AI authorship do not make sense in terms of 'the realities of the world in which the problem exists.' (Samuelson, 1985) Those realities, we argue, must push us past bare doctrinal or utilitarian considerations of originality, assessed in terms of what an author must do. Instead, what they demand is an ontological consideration of what an author must be. The ontological question, we suggest, requires an account of authorship that is relational; it necessitates a vision of authorship as a dialogic and communicative act that is inherently social, with the cultivation of selfhood and social relations as the entire point of the practice. Of course, this ontological inquiry into the plausibility of AI-authorship transcends copyright law and its particular doctrinal conundrums, going to the normative core of how law should — and should not — think about robots and AI, and their role in human relations.

Download the article from SSRN at the link.

Brkan on Freedom of Expression and AI: On Personalization, Disinformation, and (Lack Of) Horizontal Effect of the Charter @MajaBrkan

Maja Brkan, Maastricht University Faculty of Law, has published Freedom of Expression and Artificial Intelligence: On Personalisation, Disinformation and (Lack Of) Horizontal Effect of the Charter. Here is the abstract.

The question whether the use of Artificial Intelligence can interfere with the freedom of expression and information has attracted a considerable amount of attention on both sides of the Atlantic. This paper contributes to this debate by questioning whether the Charter of Fundamental Rights of the EU can be considered as an appropriate instrument to solve this tension – a perspective which has hitherto been neglected in the doctrine. The paper argues that the freedom of expression in an online environment can be endangered through the use of machine learning for personalisation, demotion of fake news (disinformation), automated blocking of illegal content and automated enforcement of the right to be forgotten. However, the application of Article 11 of the Charter to these interferences faces insurmountable obstacles due to its personal scope of application. Given that the text of this provision prevents only interferences by public authorities and thus seems to expressly exclude those by private companies such as (social) media, it is questionable whether the Charter can prevent such interferences at all. Indeed, the Charter in principle protects the citizens only against actions of public authorities (vertical direct effect) and not against other private entities (horizontal direct effect). Despite the recent recognition of horizontal direct effect of non-discrimination, effective judicial protection and the right to paid annual leave in Egenberger and Bauer, this paper argues that a similar conclusion cannot be made for freedom of expression and information. In the absence of comprehensive EU secondary legislation on freedom of expression which could compensate for such a lack of horizontal effect, citizens seem to be left without adequate protection of their rights in cases of unjustified removal of content wrongfully labelled as disinformation or illegal. This calls for further legislative action on the level of the EU.

Download the article from SSRN at the link.

Shur-Ofry and Pessach on Robotic Collective Memory

Michal Shur-Ofry and Guy Pessach, both of the Hebrew University of Jerusalem Faculty of Law, are publishing Robotic Collective Memory in the Washington University Law Review. Here is the abstract.

The various ways in which robots and AI will affect our future society are at the center of scholarly attention. This essay, conversely, concentrates on their possible impact on humanity’s past, or more accurately, on the ways societies will remember their joint past. We focus on the emerging use of technologies that combine AI, cutting-edge visualization techniques, and social robots, in order to store and communicate recollections of the past in an interactive human-like manner. We explore the use of these technologies by remembrance institutions and their potential impact on collective memory. Taking a close look at the case study of NDT (New Directions in Testimony) — a project that uses ‘virtual witnesses’ to convey memories from the Holocaust and other mass atrocities — we highlight the significant value, and the potential vulnerabilities, of this new mode of memory construction. Against this background, we propose a novel concept of memory fiduciaries that can form the basis for a policy framework for robotic collective memory. Drawing on Jack Balkin’s concept of “information fiduciaries” on the one hand, and on studies of collective memory on the other, we explain the nature of and the justifications for memory fiduciaries. We then demonstrate, in broad strokes, the potential implications of this new conceptualization for various questions pertaining to collective memory constructed by AI and robots. By so doing, this Essay aims to start a conversation on the policies that would allow algorithmic collective memory to fulfill its potential, while minimizing its social costs. On a more general level, it brings to the fore a series of important policy questions pertaining to the intersection of new technologies and inter-generational collective memory.  

Download the article from SSRN at the link.

Surden on AI and Law: An Overview and History @HarrySurden

Now available:  a lecture on Artificial Intelligence and Law, from Harry Surden, University of Colorado Law School. Watch it here. 

Chen on Judicial Analytics and the Great Transformation of American Law @Daniel_L_Chen

Daniel L. Chen, irecteur de Recherche, Centre National de la Recherche Scientifique, Toulouse School of Economics, Institute for Advanced Study in Toulouse, University of Toulouse Capitole, Toulouse, France, is publishing Judicial Analytics and the Great Transformation of American Law in the Journal of Artificial Intelligence and the Law. Here is the abstract.

Predictive judicial analytics holds the promise of increasing efficiency and fairness of law. Judicial analytics can assess extra-legal factors that influence decisions. Behavioral anomalies in judicial decision-making offer an intuitive understanding of feature relevance, which can then be used for debiasing the law. A conceptual distinction between inter-judge disparities in predictions and inter- judge disparities in prediction accuracy suggests another normatively relevant criterion with regards to fairness. Predictive analytics can also be used in the first step of causal inference, where the features employed in the first step are exogenous to the case. Machine learning thus offers an approach to assess bias in the law and evaluate theories about the potential consequences of legal change.

Download the article from SSRN at the link.

Abbas, Senges, and Howard on A Hippocratic Oath for Technologists

Ali Abbas, Neely Center for Ethical Leadership and Decision Making, Max Senges, Stanford University Center on Democracy, Development and the Rule of Law; Google, Inc; and Ronald A. Howard, Stanford Graduate School of Business, are publishing A Hippocratic Oath for Technologists in Next Generation Ethics: Engineering a Better Society (2018). Here is the abstract.

This paper presents an ethical creed, which we refer to as the Hippocratic Oath for Technologists. The creed is built on three fundamental pillars: proactively understanding the ethical implications of technology for all stakeholders, telling the truth about the capabilities, advantages, and disadvantages of a technology, and acting responsibly in situations you find morally challenging. The oath may be taken by students at Universities after understanding its basic definitions and implications, and it may also be discussed with technology firms and human resources departments to provide the necessary support and understanding for their employees who wish to abide by the norms of this oath. This work lays the foundations for the arguments and requirements of a unified movement, as well as a forum for signing up for the oath to enable its wide-spread dissemination.

Download the essay from SSRN at the link.

Brennan-Marquez and Henderson on AI and Role-Reversible Judgment @UofOklahomaLaw

Kiel Brennan-Marquez, University of Connecticut School of Law, and Stephen E. Henderson, University of Oklahoma College of Law, have published Artificial Intelligence and Role-Reversible Judgment.

As intelligent machines begin more generally outperforming human experts, why should humans remain ‘in the loop’ of decision-making? One common answer focuses on outcomes: relying on intuition and experience, humans are capable of identifying interpretive errors — sometimes disastrous errors — that elude machines. Though plausible today, this argument will wear thin as technology evolves. Here, we seek out sturdier ground: a defense of human judgment that focuses on the normative integrity of decision-making. Specifically, we propose an account of democratic equality as ‘role-reversibility.’ In a democracy, those tasked with making decisions should be susceptible, reciprocally, to the impact of decisions; there ought to be a meaningful sense in which the participants’ roles in the decisional process could always be swapped. Role-reversibility infuses the act of judgment with a ‘there but for the grace of god’ dynamic and, in doing so, casts judgment as the result of self-rule. After defending role-reversibility in concept, we show how it bears out in the paradigm case of criminal jury trials. Although it was not the historical impetus behind the jury trial — at least, not in any strong sense — we argue that role-reversibility explains some of the institution’s core features and stands among the best reasons for its preservation. Finally, for the sci-fi enthusiasts among us, role-reversibility offers a prescription as to when the legal system will be ready for robo-jurors and robo-judges: when it incorporates robo-defendants.

Download the article from SSRN at the link.

Feldman and Thieme on Competition at the Dawn of AI

Robin Feldman and Nick Thieme, both of the University of California Hastings College of the Law, are publishing Competition at the Dawn of Artificial Intelligence in the Journal of Antitrust Enforcement. Here is the abstract.

AI is in a state of perpetual and increasing revolution, with the scale of each change fully eclipsing the last. Although truly sentient machines are science fiction, experts believe society is on the verge of a technological tipping point, making future advancements unrecognizable by today’s standards. This transition may shortly resemble the shift from analog to digital technology — or from personal computing to the Internet — in which, both actively and passively, consciously and unconsciously, willingly and unwillingly, no aspect of our daily lives remains untouched. As we navigate this extraordinary gateway, what types of intellectual property policies will foster the innovation that secures our place at the head of the revolution? At the heart of this issue lies the question of how to manage forms of invention in which a computer participates, or even dominates, the inventive process. The question can be thought of as a matching problem, trying to connect the entity that invented a product, on one side, with the proper allocation of inventor rights on the other. Within the theoretical framework of the patent system, the aim of this matching problem is to incentivize innovation. Specifically, if the group of potential inventors includes both humans and computers, any invention must be either 1) created by humans alone, 2) invented in tandem by humans and computers, 3) invented by computers alone, or 4) “invented” by no neither. The paper examines each aspect of the matching equation, recommending a pathway for optimizing innovation and competition.

Download the article from SSRN at the link.

Bessen on The Policy Challenge of Artificial Intelligence @bu_law

James E. Bessen, BU School of Law, Technology & Policy Research Initiative, has published The Policy Challenge of Artificial Intelligence in the CPI Antitrust Chronicle, June 2018. Here is the abstract.

New "artificial intelligence" (AI) technology promises to bring dramatic social and economic changes, demanding major policy changes. In intellectual property and antitrust law, AI will exacerbate a damaging trend: across all major sectors of the economy, proprietary information technology is increasing the market dominance of large firms. This trend might not seem like bad news, but it is evidence of a slowdown in the spread of technical knowledge throughout the economy. The result is rising industry concentration, slower productivity growth and growing wage inequality. The key challenge to IP and antitrust policy will be counter this trend yet maintain innovation incentives.

Download the article from SSRN at the link.

Davis on Law WIthout Mind: AI, Ethics, and Jurisprudence

Joshua P. Davis, University of San Francisco School of Law, has published Law Without Mind: AI, Ethics, and Jurisprudence. Here is the abstract.

Anything we can conceive that computers may do, it seems that they end up doing and that they end up doing it better than us and much sooner than we expected. They have gone from calculating mathematics for us to creating and maintaining our social networks to serving as our personal assistants. We are told they may soon become our friends and make life and death decisions driving our cars. Perhaps they will also take over interpreting our laws. It is not that hard to conceive of computers doing so to the extent legal interpretation involves mere description or prediction. It is much harder to conceive of computers making substantive moral judgments. So the ultimate bulwark against ceding legal interpretation to computers—from having computers usurp the responsibility and authority of attorneys, citizens, and even judges—may be to recognize the role of moral judgment in saying what the law is. That possibility connects the cutting edge with the traditional. The central dispute in jurisprudence for the past half century or more has been about the role of morality in legal interpretation. Suddenly, that dispute has great currency and urgency. Jurisprudence may help us to clarify and circumscribe the role of computers in our legal system. And contemplating AI may help us to resolve jurisprudential debates that have vexed us for decades. Download the article from SSRN at the link.

Download the article from SSRN at the link.