Three Laws

Bambauer and Surdeanu on Authorbots @dbambauer @msurd @uarizonalaw @JournalSpeech

Derek E. Bambauer, University of Arizona College of Law, and Mihai Surdeanu, University of Arizona Department of Computer Science, are publishing Authorbots in volume 3 of the Journal of Free Speech Law. Here is the abstract.

ChatGPT has exploded into the popular consciousness in recent months, and the hype and concerns about the program have only grown louder with the release of GPT-4, a more powerful version of the software. Its deployment, including with applications such as Microsoft Office, has raised questions about whether the developers or distributors of code that includes ChatGPT, or similar generative pre-trained transformers, could face liability for tort claims such as defamation or false light. One important potential barrier to these claims is the immunity con-ferred by 47 U.S.C. § 230, popularly known as “Section 230.” In this Essay, we make two claims. First, Section 230 is likely to protect the creators, distributors, and hosts of online services that include ChatGPT in many cases. Users of those services, though, may be at greater legal risk than is commonly believed. Second, ChatGPT and its ilk make the analysis of the Section 230 safe harbor more com-plex, both substantively and procedurally. This is likely a negative consequence for the software’s developers and hosts, since complexity in law tends to generate uncertainty, which in turn creates cost. Nonetheless, we contend that Section 230 has more of a role to play in legal questions about ChatGPT than most commentators do—including the principal legislative drafters of Section 230—and that this result is generally a desirable one.

Download the essay from SSRN at the link.

Chesterman on From Ethics to Law: Why, When, and How to Regulate AI @ProfChesterman @nus_law @Elgar_Law

Simon Chesterman, National University of Singapore Faculty of Law, is publishing From Ethics to Law: Why, When, and How to Regulate AI in The Handbook of the Ethics of AI (David J. Gunkel, ed., Edward Elgar Publishing, forthcoming). Here is the abstract.

The past decade has seen a proliferation of guides, frameworks, and principles put forward by states, industry, inter- and non-governmental organizations to address matters of AI ethics. These diverse efforts have led to a broad consensus on what norms might govern AI. Far less energy has gone into determining how these might be implemented — or if they are even necessary. This chapter focuses on the intersection of ethics and law, in particular discussing why regulation is necessary, when regulatory changes should be made, and how it might work in practice. Two specific areas for law reform address the weaponization and victimization of AI. Regulations aimed at general AI are particularly difficult in that they confront many ‘unknown unknowns’, but the threat of uncontrollable or uncontainable AI became more widely discussed with the spread of large language models such as ChatGPT in 2023. Additionally, however, there will be a need to prohibit some conduct in which increasingly lifelike machines are the victims — comparable, perhaps, to animal cruelty laws.

Download the essay from SSRN at the link.

Loideain on Lawfulness and Police Use of Facil Recognition in the UK: Article 8 ECHR and Bridges v. South Wales Police @NoraNiLoideain

Nora Ni Loideain, Information Law & Policy Centre; University of Cambridge, Faculty of Law, Faculty of Humanities, is publishing Lawfulness and Police Use of Facial Recognition in the UK: Article 8 ECHR and Bridges v South Wales Police in Facial Recognition in the Modern State (R. Matulionyte and M. Zalnieriute, eds., Cambridge University Press 2024).

Police use of facial recognition is on the rise across Europe and beyond. Public authorities state that these powerful algorithmic systems could play a major role in assisting to prevent terrorism, reduce crime, and to safeguard vulnerable persons. There is also an international consensus that these systems pose serious risks to the rule of law and several human rights, including the right to private life, as guaranteed under the European Convention on Human Rights (ECHR). The world’s first case examining the legality of a facial recognition system deployed by police, Bridges v South Wales Police, thus remains an important precedent for policymakers, courts, and scholars worldwide. This chapter focuses on the role and influence of the right to private life, as enshrined in Article 8 ECHR, and the relevant case law of the European Court of Human Rights, in the ‘lawfulness’ assessment of the police use of live facial recognition in this landmark case. A framework which the Court of Appeal for England and Wales held was ‘not in accordance with the law’ and therefore in breach of Article 8 ECHR. The analysis also identifies and proposes new legality safeguards for police use of facial recognition. Finally, the chapter considers the importance of the emerging policy discourse prompted by Bridges in the UK surrounding the need for new legislation, a significant shift away from the current AI governance approach of combining new ethical standards with existing law.

Download the essay from SSRN at the link.

Kasdan and Pattengale on Once Thought Far-Off-In-The-Future Challenges to Copyright Law Posed by Artificial Intelligence Have Arrived: And I for One-Gulp-Welcome Our New Robot Overlords @michaelkasdan @WigginandDana @nyulaw

Michael Kasdan, NYU School of Law, and Brian A. Pattengale, Wiggin and Dana, LLP, are publishing Once Thought Far-Off-In-The-Future Challenges to Copyright Law Posed by Artificial Intelligence Have Arrived: And I for One—Gulp—Welcome Our New Robot Overlords in 58 les Nouvelles--Journal of the Licensing Executives Society (March 2023). Here is the abstract.

While we are still not in the realm of sentient machines, current-day iterations of artificial intelligence (AI) are demonstrating themselves to be potential sources of creativity which are challenging our understanding of authorship in the context of intellectual property. If AI can create new content based upon its “perceptions,” how is that distinct from human authorship and furthermore, how does or how should AI fit into the current U.S. copyright landscape?

Garon on A Practical Introduction to Generative AI, Synthetic Media, and the Messages Found in the Latest Medium @GaronJon @NSULawCollege

Jon Garon, Shepard Broad College of Law, has published A Practical Introduction to Generative AI, Synthetic Media, and the Messages Found in the Latest Medium. Here is the abstract.

OpenAI’s text generation program ChatGPT and the text-to-image generators Stable Diffusion and Dall-E have broken records for early public adoption, capital investment, and a technological shift potentially more far-reaching than even the internet itself. The broad category of generative AI has the potential to disrupt industry, art, and culture, both if done poorly and if done well. Despite significant problems with accuracy and deep concerns about the social and legal consequences of the premature adoption of these technologies, global multinational enterprises are moving these projects out of the test labs and into everyday use. This article provides a comprehensive, but introductory overview of the development of generative AI, the training methods used to produce artificially generated content, the industry opportunities for generative AI, and the legal considerations that enterprises adopting these technologies should consider. After discussing the development and implementation of the technology, the article emphasizes the key concerns regarding algorithmic bias, adherence to civil rights laws and community standards, concerns regarding defamation, and legal liability under intellectual property laws including copyright, trademark, and trade secret. The article also provides common sense steps that enterprises adopting early generative AI systems should incorporate into their agreements with the producers of generative AI networks.

Download the article from SSRN at the link.

Sarel on Restraining ChatGPT @dr_roee @unihh

Roee Sarel, Institute of Law and Economics, University of Hamburg, is publishing Restraining ChatGPT in the UC Law SF Journal. Here is the abstract.

ChatGPT is a prominent example of how Artificial Intelligence (AI) has stormed into our lives. Within a matter of weeks, this new AI—which produces coherent and human-like textual answers to questions—has managed to become an object of both admiration and anxiety. Can we trust generative AI systems, such as ChatGPT, without regulatory oversight? Designing an effective legal framework for AI requires answering three main questions: (i) is there a market failure that requires legal intervention? (ii) should AI be governed through public regulation, tort liability, or a mixture of both? and (iii) should liability be based on strict liability or a fault-based regime such as negligence? The law and economics literature offers clear considerations for these choices, focusing on the incentives of injurers and victims to take precautions, engage in efficient activity levels, and acquire information. This Article is the first to comprehensively apply these considerations to ChatGPT as a leading test case. As the United States is lagging behind in its response to the AI revolution, I focus on the recent proposals in the European Union to restrain AI systems, which apply a risk-based approach and combine regulation and liability. The analysis reveals that this approach does not map neatly onto the relevant distinctions in law and economics, such as market failures, unilateral versus bilateral care, and known versus unknown risks. Hence, the existing proposals may lead to various incentive distortions and inefficiencies. The Article, therefore, calls upon regulators to place a stronger emphasis on law and economics concepts in their design of AI policy.

Download the article from SSRN at the link.

Heller on The Concept of "The Human" in the Critique of Autonomous Weapons @kevinjonheller @MilstudiesCPH @harvardnsj

Kevin Jon Heller, University of Copenhagen, Centre for Military Studies, is publishing The Concept of 'The Human' in the Critique of Autonomous Weapons in volume 14 of the Harvard National Security Journal. Here is the abstract.

The idea that using “killer robots” in armed conflict is unacceptable because they are not human is at the heart of nearly every critique of autonomous weapons. Some of those critiques are deontological, such as the claim that the decision to use lethal force requires a combatant to suffer psychologically and risk sacrifice, which is impossible for machines. Other critiques are consequentialist, such as the claim that autonomous weapons will never be able to comply with international humanitarian law (IHL) because machines lack human understanding and the ability to feel compassion. This article challenges anthropocentric critiques of AWS. Such critiques, whether deontological or consequentialist, are uniformly based on a very specific concept of “the human” who goes to war: namely, the Enlightenment subject who perceives the world accurately, understands rationally, is impervious to negative emotions, and reliably translates thought into action. Decades of research in cognitive psychology indicate, however, that the Enlightenment subject does not exist. On the contrary, human decision-making is profoundly distorted by cognitive and social biases, negative emotions, and physiological limitations — particularly when humans find themselves in dangerous and uncertain situations like combat. Given those flaws, and in light of rapid improvement in sensor and AI technology, it is only a matter of time until autonomous weapons are able to comply with IHL better than human soldiers ever have or ever will. The article itself is divided into five sections. Section I critiques deontological objections to autonomous weapons. It shows that those objections either wrongly anthropomorphize AWS by assuming they “decide” on targets in a manner similar to humans or are predicated on a romanticized and anachronistic view of war in which most killing takes place face-to-face between combatants of equal status. Section II addresses the common argument that IHL compliance requires human understanding — particularly the ability to discern the intentions of potential targets. The section demonstrates that such understanding is far less necessary to IHL than AWS critics assume and explains why, in those situations in which judgment is necessary, limits on human decision-making undermine the idea that human soldiers are more likely to comply with IHL than autonomous weapons. The section also explain why the concept of "meaningful human control" is flawed and should give way to the superior concept of "meaningful human certification." Section III responds to the claim that autonomous weapons will not be able to comply with IHL as well as human soldiers because machines cannot feel compassion. It shows that compassion is irrelevant to IHL compliance, that compassion can lead to negative outcomes in combat as well as positive ones, and that any potential benefits of compassion are far outweighed by the costs of negative emotions such as stress and anger. Section IV addresses the argument that the non-human nature of autonomous weapons makes it difficult, if not impossible, to hold humans responsible for war crimes that AWS commit. The section demonstrates not only that the problem of “accountability gaps” is significantly overstated, but also that there is no significant difference between human soldiers and autonomous weapons in terms of criminal responsibility. Finally, Section V explores the implications of the idea that it is highly likely autonomous weapons will eventually be able to comply with IHL as well as — if not better than — human soldiers. It argues that consequentialist critics are not primarily concerned AWS will be worse soldiers than humans. Instead, their real worry is that they will be better ones, because the more humane war becomes, the more difficult it will be to eliminate war itself. This, the section argues, is actually the most powerful argument against autonomous weapons — but one that applies to most of the weapons developed over the past century.

Download the article from SSRN at the link.

Ringel and Reid on Regulating Facial Recognition Technology: A Taxonomy of Regulatory Schemata @evanringel3 @UNC

Evan Ringel and Amanda Reid, both of the University of North Carolina, Chapel Hill, have published Regulating Facial Recognition Technology: A Taxonomy of Regulatory Schemata and First Amendment Challenges at Communication Law and Policy, 2023, AHEAD-OF-PRINT at 1. Here is the abstract.

This article examines the patchwork of regulatory approaches policymakers have used to govern use of facial recognition technology (FRT). Without comprehensive federal legislation, state and local policymakers are left to fill the regulatory gap. The purpose of this article is two-fold. First, it creates a taxonomy of the existing regulatory schemata governing uses of FRT. The authors’ systematic analysis of FRT regulatory approaches identified five main categories of policymaking options: (1) Agents of Use; (2) Limitations on Use; (3) Accountability of Use; (4) Evaluation of Use; and (5) Enforcement of Permitted Use. Second, the authors analyze whether First Amendment protection of information as speech may serve as a barrier to any regulatory aspect of FRT. Building on the novel empirical framework of regulatory options, the authors examine which types of FRT regulations would likely survive a First Amendment challenge. Thus, this taxonomy of regulatory approaches, coupled with a First Amendment analysis, offers valuable insights for policymakers and scholars.

Download the article from SSRN at the link.

Romig on The Ethics of ChatGPT: A Legal Writing and Ethics Professor's Perspective @JenniferMRomig @EmoryLaw

Jennifer Murphy Romig, Emory University School of Law, has published The Ethics of ChatGPT: A Legal Writing and Ethics Professor’s Perspective as Emory Legal Studies Research Paper. Here is the abstract.

Teaching law students requires meeting them where they are, envisioning what they can become, and offering appropriate education and advice for their developmental journey. ChatGPT is not a person, but this essay will to some extent personify ChatGPT in discussing how it might be developed before it is used for client work. ChatGPT and A.I. text generators don’t “learn” but they can be programmed, prompted, and trained. Everyone realizes their potential to become pervasive and ubiquitous in the legal field and, indeed, in every field that relies on the production and exchange of words. But pervasive and ubiquitous does not necessarily mean ethical or effective. So far in using ChatGPT I have seen a broad range of embarrassingly wrong output and really promising output. My comments here are offered in the spirit of what I’d like to see from ChatGPT.

Download the paper from SSRN at the link.

Delfino on The Deepfake Defense: Exploring the Limits of the Law and Ethical Norms in Protecting Legal Proceedings from Lying Lawyers @RebeccaDelfino @LoyolaLawSchool

Rebecca Delfino, Loyola Law School, Los Angeles, has published The Deepfake Defense—Exploring the Limits of the Law and Ethical Norms in Protecting Legal Proceedings from Lying Lawyers as Loyola Law School, Los Angeles Legal Studies Research Paper No. 2023-02.

Thousands of audiovisual images documented the insurrectionists who stormed the United States Capitol on January 6, 2021. Authorities subsequently collected those images and charged some for their criminal conduct. Given the overwhelming audiovisual evidence implicating the insurrectionists, it should be impossible to assert a plausible defense claiming that those unmistakably depicted in the images were not present. Right? Wrong. As the defense in the federal criminal trial of January 6th insurrectionist leader Guy Reffitt illustrated, the emergence of “deepfakes” has changed the landscape of plausible defenses to crimes. Reffitt led the attack on the Capital. Videos and other visual images showed him at the head of the crowd advancing on the Capitol’s West Terrace. He was arrested and charged with multiple crimes. And although the evidence, including audiovisual images, against Reffitt, was clear and overwhelming, his lawyer undermined it, arguing to the jury that the evidence against Reffitt was a “deepfake” – an audiovisual recording created using Artificial Intelligence technology that allows anyone with a smartphone to believably map one person’s movements and words onto the image of another person. Unfortunately, the law does not provide a clear response to Reffitt’s lawyer’s reliance on deepfakes as a defense. But this much is clear—the “deepfake defense” is a new challenge to our legal system’s adversarial process and truth-seeking function. Because the norms of professional ethics require lawyers to advocate zealously, deepfakes invite lawyers to raise objections and arguments to evidence to exploit juror bias and skepticism about what is real. Thus, lawyers may plant the seeds of doubt in jurors’ minds to question the authenticity of all digital audio and visual images, even those counsel knows to be genuine. Currently, no rule of procedure, ethics, or legal precedent directly addresses the presentation of the “deepfake defense” in court. The existing standards provide scant guidance because they were developed before the advent of deepfake technology. As a result, they do not solve the concern of how to deter lawyers from exploiting it. Although in the last several years, legal scholarship and the popular news media have addressed certain facets of deepfakes, there has been no in-depth commentary on the “deepfake defense.” This article is the first to explore the deepfake defense, locating it within the historical and current framework of lawyers’ efforts to fabricate evidence and the laws and the practice norms that exist to curb that conduct. It proposes a reconsideration of the ethical rules governing candor, fairness, and the limits of zealous advocacy and urges a re-examination of the court’s role in sanctioning such conduct. Thus, this article offers novel proposals to guide the way forward for lawyers and courts as they traverse this new technological landscape.

Download the article from SSRN at the link.