29 July 2023

Animal Personhoods

'Evolving Conceptions of Legal Personhood: What Might Recent Legal Developments Herald for Non-Human Animals in Australia' by Jane Kotzmann, Morgan Stonebridge and Paulien Christiansen in (2023) 46(2) UNSW Law Journal 347 comments 

 Legal personhood has traditionally been associated with human persons and their representatives, for example, corporations. Recent years, however, have seen the binary conception of legal personhood challenged and reconceived, and the circle of legal persons expanded in numerous jurisdictions. In particular, the utter failure of environmental laws to protect the environment has led to the recognition of nature bodies as legal persons. Within this context, this article considers whether these developments might lead to recognition of animals as legal persons in Australia. The parallel deficiencies of environmental and animal laws, together with the willingness of some Australian legislatures to entertain legal personhood for nature, may suggest that the concept of legal personhood for animals in Australia is not completely far-fetched.  ... 

Maitland ... described a legal ‘person’ as ‘a right-and-duty-bearing unit’. Implicit in that description ... is the traditional, essentially functional, understanding of legal ‘personality’ as lying in the existence of legally conferred or legally recognised capacity to have or to form legal relations. Implicit also is the traditional understanding of legal personality as unitary. 

The legal landscape appears to be shifting from the traditional view of legal personhood described by the High Court in the above quote. This traditional view sees legal personhood as being based on individual autonomy and rationality: a legal person should be able to independently exercise their legal rights and carry out legal duties. The traditional approach has, however, been subject to significant criticism. It has been attacked on the basis that it fails to protect the interests of vulnerable individuals that lack the capacity to exercise legal rights, including animals, and that it ignores the interdependence of individual persons, animals and the environment. Further, it has been argued that the concept itself is inexorably tied to ‘a colonial and otherwise exclusionary logic’.Accordingly, scholars have sought to reformulate personhood – for example, as ‘relational personhood’ – or have rejected the concept and advocated for adoption of an alternative legal model that better recognises vulnerability and interconnectedness. 

In tandem with these conceptual developments, the circle of legal persons has been expanding. Outside of human persons, the law has long recognised their representatives – for example, corporations or churches – as legal persons. Laws in some jurisdictions have also recognised deities, or gods, as legal persons. More radically, and largely as a response to the overwhelming failures of environmental law to protect the environment, legislatures and courts have been engaging in ‘legal experimentation’ in recognising bodies of nature as legal persons. For example, the Whanganui River in New Zealand, all rivers in Bangladesh9 and the Mar Menor lagoon in Spain have all been declared legal persons. A small number of jurisdictions have gone even further and recognised animals as legal persons. 

In the context of sustained advocacy for legal personhood for animals and research indicating very high levels of public concern for animals in Australia, this article investigates whether these developments in relation to legal personhood are likely to have any implications for non-human animals in Australia. In this respect, the article seeks to ascertain whether there is any reasonable prospect of animal legal personhood in Australia but remains largely neutral in terms of whether advocates should seek the attribution of personhood (perhaps with a view of progressing to an improved legal model in the future) or of some alternative legal model. It assumes that an improved legal status for animals is necessary and desirable, but does not analyse the various merits of legal personhood and its proposed replacements. 

The willingness of some Australian legislatures to adapt legal personhood models for nature bodies may suggest that the concept of animals as legal persons is not as far-fetched as it might seem. While existing research has considered the potential for animal legal personhood in Australia,  this article analyses the conceptual challenges to legal personhood and recent legislative changes to determine the practicability of arguing for legal personhood for animals in Australia. Further, in the event of legal personhood recognition for animals, this article draws on the experience of nature rights to identify considerations that should influence personhood construction for animals. 

This article contends that the evolution of the concept of legal personhood over recent decades, public concern for animals, legal recognition of animal sentience and legal experimentation in relation to nature rights in Australia, suggests that personhood may be a realistic path to greater legal protection for animals in the medium to long-term future in Australia. The article focuses on Australia because a thorough discussion of the potential for global personhood developments would require a lengthier analysis, beyond the word limitations of a typical journal article. Australia also presents a strong case study given that it has a comparatively high rate of meat consumption and is not considered a leader in matters of animal protection. As such, the potential for change found in Australia is likely also applicable to other jurisdictions and would potentially meet less resistance. In terms of implementation, however, thought should be given to how legal personhood would best be framed for animals. In this respect, attention should be paid to who should be enabled to speak on behalf of animals and what powers they should be given. While legal personhood for animals would constitute a profound legal achievement, it would nevertheless be insufficient to challenge the broader economic and political frameworks that rely on animal exploitation. As Benjamin J Richardson and Nina Hamaski identify in relation to nature rights, ‘[p]roperty tenure, markets, business corporations, economic growth policies and other drivers of environmental upheaval remain intact’. The same critique may be made in relation to potential legal personhood for animals. Nevertheless, along with other legal developments, legal personhood may assist in shifting society forward to a point where challenges directed at political and economic frameworks can be made. 

The next Part of this article explores the meaning and significance of legal personhood. This includes analysis of why the concept is important in the context of protecting individual interests, how it has traditionally been conceived, challenges to those traditional conceptions and the connection between legal personhood and rights. This discussion serves to highlight both the importance of legal personhood as well as its current fluid state. Part III outlines the ways in which the circle of legal persons has been expanding in recent years as part of pressing ‘legal experimentation’.  Part IV then addresses some common criticisms of advocacy for animal legal personhood.

26 July 2023

Compliance

'Compliance Gatekeepers' by Roy Shapira and Asaf Eckstein in Yale Journal on Regulation comments 

What determines the effectiveness of corporate compliance programs, and who is accountable when they fail? Scholars and policymakers tend to answer these questions by focusing on internal compliance actors: directors, CEOs, general counsels, chief financial officers, and chief compliance officers. Yet in reality, all these corporate insiders rarely perform compliance tasks on their own. They rather heavily rely on outside compliance advisors. In this Article we spotlight the understudied role of outside compliance advisors and make the following three contributions. 

First, we document the various functions that outside compliance gatekeepers play these days: from designing reporting systems, to conducting internal investigations and monitorships, to performing “racial equity audits” and verifying ESG disclosures. Along all these dimensions, there exists a gap between the high level of expectations for compliance gatekeepers to improve corporate behavior and their low levels of accountability for compliance failures. 

Second, we examine the causes of compliance gatekeepers’ lack of accountability. Compliance gatekeepers rarely face litigation, even after colossal compliance failures, because an amalgamation of doctrines set a very high pleading hurdle across all potential claims against them. Private ordering is ineffective too. The buyers in the market for compliance gatekeeping – namely, corporate insiders – do not necessarily want outside compliance gatekeepers to hinder their company from making profits by skirting regulations in real time. Nor do corporate insiders want outside gatekeepers to probe their internal affairs diligently after the fact and place the blame for corporate wrongdoing at their feet. By paying outside compliance gatekeepers with shareholders’ money, corporate insiders buy plausible deniability for themselves. From the sellers’ (gatekeepers’) perspective, it is convenient to ramp up expectations while not being too stringent with their clients, because this keeps a newfound revenue stream alive and growing. The gap between high expectations and somber reality may therefore be a feature rather than a bug. 

Finally, we propose concrete policy measures that could improve corporate compliance. For example, public enforcers need to rethink the practice of providing lenient treatment to corporate wrongdoers who rely on outside experts. Credit to wrongdoers should be conditioned on outside experts facing a meaningful threat of liability or at minimum transparency. Courts need to rethink the applications of doctrines such as “in pari delicto,” which blocks claims of professional negligence and breach of contract against gatekeepers. And corporate law courts should interpret shareholders’ right to inspect their company’s books more liberally, so that it includes access to documents pertaining to gatekeepers’ work, thereby enabling shareholders to investigate potential gatekeeper misconduct.

Cheating

'Widespread use of Chegg for academic misconduct: Perspective from an Australian engineering school' by Edmund Pickering and Clancy Schuller comments 

Online tools are increasingly being used by students to cheat. File-sharing and homework-helper websites offer to aid students in their studies, but are vulnerable to misuse, and are increasingly reported as a major source of academic misconduct. Chegg.com is the largest such website. Despite this, there is little public information about the use of Chegg as a cheating tool. This is a critical omission, as for institutions to effectively tackle this threat, they must have a sophisticated understanding of their use. To address this gap, this work reports on a comprehensive audit of Chegg usage conducted within an Australian university engineering school. We provide a detailed analysis of the growth of Chegg, its use within an Australian university engineering school, and the wait time to receive solutions. Alarmingly, we find Chegg is broadly used to cheat and 50% of questions asked on Chegg are answered within 1.5 hours. This makes Chegg an appealing tool for academic misconduct in both assignment tasks and online exams. This work provides valuable insights to educators and institutions looking to improve the integrity of their courses through assessment and policy development. Finally, to better understand and tackle this form of misconduct, we call on education institutions to be more transparent in reporting misconduct data and for homework-helper websites to improve defences against misuse. 

Academic integrity is of critical importance to the modern tertiary education sector and underpins pedagogical approaches to teaching and learning. To ensure the integrity of their courses, academics and institutions must be aware of the latest methods students use to cheat. Universities must be agile in their response to new trends in academic integrity and student misconduct, especially given the new challenges and opportunities offered by the digital era. While early academic integrity research placed emphasis on plagiarism (Walker, 1998), new forms of misconduct are growing, including homework-helper, contract cheating and file- sharing websites (Curtis et al., 2022; Lancaster & Cotarlan, 2021a), automatic text-spinning and paraphrasing tools (Prentice & Kinden, 2018; Rogerson & McCarthy, 2017) and AI tools (Finnie-Ansley et al., 2022). Of interest in this paper is the growing prevalence of homework- helper websites, in particular Chegg.com (henceforth referred to as Chegg), the largest such website (Chegg Inc, 2022). 

Homework-helper websites, offer to aid students in their learning, however they also represent a growing threat to academic integrity. In this article we draw a distinction between contract-cheating websites (e.g. essay mills) which exist exclusively for the purpose of cheating, and homework-helper websites which present as a legitimate service, but whose business model is extremely vulnerable to misuse. We also make a distinction between homework-helper and filesharing websites. The latter term is commonly used in literature, but is not preferred by the authors of this article as file sharing is only one such service offered by homework-helper websites (Lancaster & Cotarlan, 2021a). 

Homework-helper websites offer to aid students in their studies through a range of services, for example question & answer (Q&A) services, file-sharing (e.g. study notes and assessment) and citation assistance. Of concern in this paper are Q&A services, which are identified as a concerning source of cheating (Broemer & Recktenwald, 2021; Lancaster & Cotarlan, 2021a). Through these Q&A services, students can look-up solutions to questions on a database or submit their own questions to be solved by the website’s ‘tutors’ (these are then added to the database). By far the largest homework-helper website is Chegg, with a market capital of 3.7 billion USD and 7.3 million subscribers (Chegg Inc, 2022; Nasdaq Inc, 2022). Other large homework-helper sites include CourseHero, Studocu and Bartleby. These websites generally operate under a subscription model by which students pay a monthly fee (14.95 USD for Chegg as of writing) to access the solution database and to ask their own questions. While these sites purport to be for legitimate study, they are highly vulnerable to misuse, and there are limited mechanisms to prevent students using these Q&A services to cheat. Broemer and Recktenwald (2021) proposed that the Chegg’s Q&A service is primarily used for cheating, this is further supported by Lancaster & Cotarlan (2021b). Figure 1 provides context on use of these sites by showing (a) a unique assessment question, (b) Google search results identifying solutions to the questions on (c-d) Chegg and (e) CourseHero. ... 

Considering a student’s motivations to cheat, homework-helper websites are highly appealing. A small portion of students routinely cheat, while a much larger group, approximately 44%, fall within the cheat-curious category – these are students who may cheat under certain circumstances (Bretag et al., 2018, 2019; Rigby et al., 2015). Drivers of cheating behaviour include a perception of low risk (Diekhoff et al., 1999), perception that there are lots of opportunities to cheat (Bretag et al., 2019) and perception of norms (Curtis et al., 2018) (i.e., the perception that cheating is the norm, or that others are cheating). 

Homework-helper websites embody these motivators. Homework-helper websites appear high in search engine results pages (often as the first result), and thus may appear when students engage in normal and healthy study behaviour (e.g. researching an assignment problem). It takes just one student to upload an item to a homework-helper website for it to appear in common search engines like Google. The appearance of these links provides a low barrier to entry, provides ample opportunity and feeds into a low perception of risk. Further, the fact that a peer has uploaded the assessment to a homework-helper site feeds into the perception that others are engaging in this behaviour. Combined, cheat curious students are vulnerable to these motivators. Finally, due to the large subscriber base and ease of uploading questions, these websites can rapidly include unique assignment questions (Christodoulou, 2022). 

The global COVID-19 pandemic has increased rates of academic misconduct and usage of homework-helper websites (Comas-Forgas et al., 2021; Erguvan, 2021; Lancaster & Cotarlan, 2021a). Subsequently, the pandemic has seen an increased focus from tertiary education institutions on modern trends in academic misconduct (Erguvan, 2021; Reedy et al., 2021; Turner et al., 2022). A recent study by Lancaster and Cotarlan (2021a) studied the impact of COVID-19 on Chegg usage, finding 196% increase following the transition to online learning. This trend is particularly concerning for engineering and other STEM disciplines. STEM is known to be overrepresented in academic misconduct (Bretag et al., 2019; Lancaster & Cotarlan, 2021a). For example, a recent Australian based large-scale survey of students found engineering students were 1.8x more likely to engage in cheating behaviour (Bretag et al., 2019). Adding to this concern, Chegg’s own data shows that a majority (59%) of their userbase are STEM students (Chegg Inc, 2022). 

While there is growing concern about homework-helper websites in the tertiary education sector, there remains little information on the use of these websites. To demonstrate, as of the date of writing, a Scopus search for ‘Chegg,’ the largest homework-helper website, returns nine journal articles, with only four related to academic misconduct, the others relating to Chegg’s textbook hiring service or geology (Busch, 2017; Emery-Wetherell & Wang, 2023; Lancaster & Cotarlan, 2021a; Ruggieri, 2020). Lancaster and Cotarlan (2021a) studied the impact of COVID-19 on Chegg usage; Ruggieri (2020) found 38 – 71% of physics students reported Chegg usage, with increased usage in more advanced units; while Busch (2017) explored methods to reduce usage of sites like Chegg; Emery-Weatherell & Wang (2023) explored cheating via Chegg in an introductory statistics course, explored ways to discourage cheating, and provided code to help identify cheating students. 

Broader searches uncover additional literature. In a valuable conference paper, Broemer and Recktenwald (2021) presented a detailed investigation of Chegg usage in a 2-hour online mechanical engineering exam, identifying 129 unique posts to Chegg, with 71% of posts answered during the exam (50% answered within 1 hour). Interestingly, many answered posts were not viewed during the exam, even by the uploader. Manoharan & Speidel (2020) uploaded assignment questions to Chegg investigating factors like solution quality and time for a solution to be provided, finding easy questions were answered quickly and correctly, while complex questions were not answered or received substandard answers. Interestingly, Manoharan & Speidel (2020), ensured the questions they uploaded were clearly identifiable as formative assessment, which violates Chegg’s policy and which Chegg tutors are required to report. None of these questions were identified as attempts to cheat. Finally, Somers et al (2023) developed a tool to automatically detect the upload of questions to the file-sharing and homework-helper sites like Chegg. 

As Chegg is the largest homework-helper website, this lack of investigation represents a major gap in our understanding of academic misconduct and cheating. With such sparse information, developing effective strategies and policies is near impossible. To address this gap, this article aims to present a detailed understanding of Chegg usage within an Australian technology university engineering school. This article provides insights into the growth of Chegg’s homework-helper service, prevalence of Chegg usage, and time required to receive a solution on Chegg. The findings provide a valuable resource for institutions and academics in understanding the challenge presented by homework-helper websites.

Licensing

'Compulsory Licensing of Intellectual Property' by Enrico Bonadio in Cristiana Sappa (ed), Research Handbook on Intellectual Property and Inclusivity (Elgar, 2024) comments 

A specific IP built-in limitation to IP rights is compulsory licencing. Compulsory licences are mostly known in patent and to a lesser extent plant variety and copyright laws, while they are extremely rare in trademark law. The essence of an IP compulsory licence is that the authorisation to exploit the protected asset (e.g., an invention, a plant variety, a copyrighted work or a brand) does not come from the right holder. It comes from the government or the judiciary instead. 

The chapter focuses on this IP flexibility and notes how this legal and policy tool re-aims at introducing elements of inclusivity into IP regimes. It argues that compulsory licences, if granted in line with the relevant requirements under international and national laws, cannot be considered a negation of IP rights. Indeed, exceptions to IP are part of the very IP regimes which grant exclusive rights in the first place: in other words, they constitute the other side of the same coin.

Regulation

'The right to regulate and the interpretation of the WTO Agreement ' by Andrew D Mitchell in (2023) Journal of International Economic Law comments 

International investment treaty arbitrators have long recognized the existence of a State’s right to regulate under international law—independent of the treaty text itself—that ought to be factored into their analysis of whether a State violated its treaty obligations. The precise contours of such a right, and its interaction with the substantive disciplines of the treaty in question, have varied considerably across international investment law disputes. Indeed, some adjudicators have taken the approach that, although such a right absolves a State of acting unlawfully under the treaty, it does not absolve the State from paying compensation under it, perhaps nullifying the right’s practical import. Nonetheless, this represents a minority view, and perhaps in response to those approaches, States have increasingly codified the proposition in the text of their international investment treaties that there exists a sovereign right to regulate to be factored into adjudicators’ interpretation of their substantive disciplines. Recent major non-World Trade Organization (WTO) plurilateral trade agreements have also explicitly recognized the existence of an inherent ‘right to regulate’ in the public interest. 

Compared to international investment law arbitrators, WTO adjudicators confronted with arguments about an inherent right to regulate that applies to WTO Agreement obligations have typically been dismissive. For instance, in China—Raw Materials, China argued that it possessed a right to regulate trade as an ‘inherent right’, and ‘not [as] a “right bestowed by international treaties such as the WTO Agreement”’, and that this inherent right should inform adjudicators’ understanding of China’s substantive commitments under the WTO Agreement. The USA, Mexico and European Union argued that China had exercised its inherent right to regulate trade when it acceded to the WTO Agreement, and any ability to deviate from its substantive disciplines can be found only in the WTO Agreement itself. Both the panel and the Appellate Body (‘AB’) agreed with the USA, Mexico and the European Union and declined to consider the existence of an inherent right to regulate outside of the WTO Agreement. 

Along these lines, WTO adjudicators have traditionally framed a State’s right to regulate as essentially residual in character. In other words, upon accession to the WTO Agreement, whatever inherent right to regulate States have had under international law is transmuted into deference for their regulatory autonomy under WTO disciplines, rather than remaining an inherent WTO right as such. A key consideration in this regard has been the existence of standalone exceptions clauses that provide policy-based exceptions to substantive disciplines in major instruments of the WTO Agreement, such as Article XX of the General Agreement on Tariffs and Trade (‘GATT’)10 and Article XIV of the General Agreement on Trade in Services (‘GATS’). Moreover, WTO Agreement drafters have explicitly cross-applied these policy-based exceptions to some other WTO instruments, such as the Agreement on Trade-Related Investment Measures. 

The existence of clauses providing for policy-based exceptions has led WTO adjudicators to eschew policy-based considerations in their interpretations of substantive disciplines, such as non-discrimination articles, even where there might be a textual or contextual basis for doing so. For example, the fourth preambular recital of the GATS ‘[r]ecogniz[es] the right of Members to regulate, and to introduce new regulations, on the supply of services within their territories…to meet national policy objectives’. The AB in Argentina – Financial Services nonetheless found this recital irrelevant to the interpretation of non-discrimination provisions in the GATS, given the codification of a more limited set of policy-based considerations through the exceptions clause of Article XIV. For the AB, the existence of policy-based exceptions meant that ‘regulatory concerns’ were excluded from the substantive disciplines. The AB reached the same conclusions under the GATT in EC—Seal Products. 

Thus, in characterizing a State’s right to regulate as essentially residual in character, WTO adjudicators have traditionally emphasized that a State must either find a basis in the text of an instrument under the WTO Agreement that explicitly justifies a regulatory action16 or pursue only those regulatory actions that are consistent with the substantive disciplines of the WTO Agreement. Accordingly, unlike in international investment arbitrations, a State’s right to regulate has not been viewed as a standalone component under the interpretive analysis in WTO dispute settlement. Instead, policy-based considerations have been integrated only to the extent that they find expression in the ordinary sources of treaty interpretation, namely, the provision’s plain text and context in light of the treaty’s object and purpose. 

This general approach has presented challenges concerning instruments under the WTO Agreement (‘Covered Agreements’) that omit any policy-based exception clauses. On occasion, WTO adjudicators have sought to overcome such challenges by cross-applying a policy-based exception clause to another instrument with a clear textual basis for doing so,19 but they have generally avoided this approach. 

This article explores how WTO adjudicators have engaged in the interpretive process in cases involving a public policy-based measure when no policy-based exception clause has been available. This question has arisen recently in a diverse array of cases under the Agreement on Subsidies and Countervailing Measures (‘SCM Agreement’) (section II), the Agreement on Trade-Related Aspects of Intellectual Property Rights (‘TRIPS Agreement’) (section III) and the Agreement on Technical Barriers to Trade (‘TBT Agreement’) (section IV). As I will show, WTO adjudicators have proactively made selective interpretive choices to recognize States’ right to regulate under these instruments, even where this may seem contrary to their plain text and the orthodox approach described earlier. While I do not take issue with the ultimate outcomes reached by WTO adjudicators, their attempts to situate a State’s right to regulate within the ordinary tools of treaty interpretation have resulted in approaches that are, at first glance, unclear and uncertain. Nevertheless, I argue in The right to regulate as a principle of regulatory autonomy? section that, while such unorthodox approaches appear ungrounded in ordinary interpretive principles, they can be explained in terms of each WTO instrument’s unique text, structure and purpose. By making more explicit reference to these legislative characteristics in their analyses, WTO adjudicators might be able to acknowledge a right to regulate at the WTO explicitly—not as an inherent and unfettered right at international law—but as a principle of regulatory autonomy that is derived explicitly or implicitly from the text of the Covered Agreements. This characterization would provide a firmer basis for WTO bodies to either reinforce or peal back policy-based considerations on a case-by-case basis.

25 July 2023

Safe and Responsible AI

'Regulation of (generative) AI requires continuous oversight' - an AustLII submission by Graham Greenleaf, Andrew Mowbray and Philip Chung to the Department of Industry, Science and Resources in response to Discussion Paper 'Safe and Responsible AI in Australia' comments

Australia’s Department of Industry, Science and Resources invited interested parties to make submissions on a Discussion Paper (DP) Safe and Responsible AI in Australia. This submission by researchers from the Australasian Legal Information Institute (AustLII) addresses the most important general issues identified in the Discussion Paper and suggests the best strategies to address them. 

In the paper, we make the following submission:

1 The rate of take-up of automated decision-making (ADM) systems in Australia, to identify those that may pose considerable risk to Australia, should be examined at the outset of determining an AI policy for Australia, and kept under regular review. 

2 Regulation should not be aimed at ‘AI generally’, but should be aimed at two things: (i) Regulation of the use of specific applications of underlying AI technologies; and (ii) Regulation, by imposition of conditions on any use of a particular underlying AI technology, and therefore (for practical purposes) of their development; but should not aim to prevent research into the technology or its application 

3 The definition of ‘AI’ should be altered so that it refers to ‘without explicit programming, or which achieves a similar result by other means’. ‘Hallucinations should be dropped, and ‘deceptive and reckless mis-statements’ or ‘fabrications’ used instead. 

4 There should be a set of principles used to guide Australian regulation, that the principles should be based on international consensus, should be as consistent as possible across all Australian jurisdictions, and should be as comprehensive as needed. 

5 We recommend ‘Ten guiding principles’, a modification of the ‘Australian principles’. The NSW Framework’s approach to implementation should be taken into account. 

6 For Australia to be able to safely and responsibly regulate AI (and particularly generative AI), there needs to be a continuous source of expert advice which will regularly report to existing regulatory bodies, to government and to the public, updating them on whether there are significant changes to our ability to uphold the principles on which regulation of AI is based, and make proposals concerning changes needed. 

7 An ‘Australian Advisory Board on Regulation of AI’ (the AI Board) would have a remit of two to three years, independence so that it could give frank advice, and an obligation to produce six monthly reports. It should preferably consist of ten members or fewer. 

8 Australia should aim to provide inputs to influence international developments where possible. In the short term, the most advanced international source of AI regulation is likely to be the European Union. In the longer term, there may be steps toward a binding international agreement concerning AI (or generative AI at least). 

9 A risk-based approach to regulation of AI should be adopted and implemented in Australia by Commonwealth legislation and regulations. Only the most dangerous applications should be brought within the regulatory structure in the first instance. 

10 The Commonwealth’s initial ‘AI Framework Act’ should involve at least the following: a. Create an Australian AI Board. b. Require assessment by government of the take-up of AI in Australia. c. Include the Ten Guiding Principles for (Generative) AI we recommend. d. Implement a risk-based approach to regulation. e. Make transparency mandatory for all AI applications impacting upon Australian individuals and organisations.

23 July 2023

LLMs

'Do Androids Dream of Electric Copyright? Comparative Analysis of Originality in Artificial Intelligence Generated Works' by AndrĂ©s Guadamuz  in (2017) 2  Intellectual Property Quarterly and updated in 2020 comments 

The advent of sophisticated artificial neural networks has opened new artistic opportunities, but also a variety of new legal challenges. Computer programs such as Google's Deep Dream can take an image and process it in manners that resemble biological networks, producing artwork that is both unique and unpredictable. 

The law is not unfamiliar with the challenges of artificial intelligence, in the past academics and policymakers have had to deal with the legal implications of autonomous agents in contract formation, just to name one are of interest. However, for the most part the implementation of smart systems has been limited in their reach and scope, and in many instances autonomous agents required quite a lot of direction from the programmer, following a very stringent set of rules. This meant that for the most part all rights, responsibilities and liabilities arising from artificial agents fell squarely on the program creator. Neural networks are different, these systems have the potential to generate works in which human interaction is minimal. 

Modern copyright law has been drafted to consider originality as an embodiment of the author’s personality, and originality is one of the main requirements for the subsistence of copyright. So, what happens when you remove personality from the equation? Are machine-created works devoid of copyright? Do we need to change copyright law to accommodate autonomous artists? 

'A Scanner Darkly: Copyright Liability and Exceptions in Artificial Intelligence Inputs and Outputs' by Guadamuz in 2023 comments 

 The question of artificial intelligence and copyright has begun to gain considerable momentum over the last few years. One area of study focuses on the authorship of computer-generated works, but arguably, a more intriguing inquiry centers around infringement. To paint, compose music, or write, AI must be taught. The process by which artificial intelligence "learns" to perform these tasks, particularly to generate works that emulate human creativity, often relies on accessing and analysing a large number of existing works to discern patterns and create its own versions. To accomplish this, the computer program needs copies of these works for analysis and subsequent generation of new outputs. 

Thus, two significant copyright questions emerge: one concerning the inputs and another the outputs. From the inputs perspective, does the act of accessing, reading, analysing, and mining data constitute copyright infringement? If so, are there any applicable defences? From the outputs perspective, could the copyright owner of one of the works used to train the computer sue the program's creator for copyright infringement due to the resulting derivative works? This article delves into both questions.