Intellectual property law and generative artificial intelligence: fair remuneration, equality or ‘My plentie makes me poore’

Abstract

• Intense debates are occurring on the regulation of artificial intelligence (AI). Most of these debates are horizontally fragmented along existing categories of legal fields, such as human rights, international trade, competition law or IP law. They are also vertically divided along the distinct lines of national, regional and global levels of lawmakers. Combined with several misconceptions about the underlying concepts, the result of this fragmentation is that it obscures the fundamental problems that need to be addressed to ensure global sustainable development in the future.

• Exemplified by the debate on the impact of AI on IP law, this article argues that the primary focus of the debate on the IP protection of AI-generated works, the liability resulting from the infringement of IP-protected works by AI-generated works and the infringement of existing IP-protected works based on the training of AI foundational models are misguided. Instead, the central focus should be on the increase in inequalities throughout the world and critically assessing the role of IP law in this trend.

• To this end, the article seeks to illustrate the shortcomings of IP law prior to the advent of AI and to re-establish the goal of the original function of IP, which is to provide fair remuneration to all types of creators to serve the benefits of wider society. It concludes that IP law’s pioneering role in innovation should continue to contribute to the establishment of a fairer system of financial compensation for all creative work in the future.

1. Introduction

My plentie makes me poore.¹

Against the backdrop of vigorous global debates about artificial intelligence (AI), it is a fundamental mistake to believe that the recent rise in AI services is driving copyright and all of IP law into a serious crisis.² It is accurate to state that OpenAI’s introduction of Chat-GPT in November 2022 and its hype took many by surprise when it registered more than 100 million active users in less than 2 months, which helped make it ‘the fastest-growing consumer technology in history’.³ It is also accurate to state that the constantly evolving abilities of these technologies are posing a challenge to IP law, especially their rapid evolution from mere chatbots to multimodal large language models, with the looming prospect of an emerging artificial general intelligence.⁴ These technologies are portrayed as capable of engendering creative content or, at least, of mimicking the human generation of creative content at unprecedented scales.⁵ However, the legal challenges posed by AI are not limited to IP law but rather extend to all areas of law or, more likely, law in its entirety. By segmenting the debates and creating a hype about AI’s possible technical capabilities, other social risks are missed. For instance, in the context of IP law, the increasing levels of inequality caused by the exploitative mechanisms, profit-oriented services and general tendencies of existing businesses to ‘accumulate capital through the rapacious extraction of user data and unremunerated labor’ are usually overlooked or deliberately masked.⁶

This article argues that the debates focused, in general, on the regulation of AI and, in particular, on possible amendments to IP law are largely misconstrued and based on a number of misconceptions. In its essence, this article uses IP law as an example to advocate for the need to address the problem of increasing levels of global inequality. It argues that the three principal questions about AI discussed in IP law, namely, the IP protection of AI-generated works, the liability for the infringement of IP-protected works by AI-generated works and the infringement of existing IP-protected works based on the training of AI foundational models, are of secondary importance and are too narrowly conceived.⁷ Instead, creative thinking is required about the future of work altogether, with the primary focus given to finding different ways to contribute to a future global economic system that ensures fair remuneration for creators and labourers of all types in order to reverse the current trends of increasing inequalities.

To correct these misconceptions in the AI-related debates, Section 2 begins with an overview of the IP-related AI governance debate and some of the flawed assumptions upon which it is based. It seeks to challenge the assumption that IP law worked well before the rise of AI as well as the myth of the poor but creative artist. Section 3 puts forward the argument that the IP crisis started long before the recent advent of AI systems by providing evidence of the failure of the IP law to provide fair remuneration and, thereby, reduce the already unacceptable levels of inequalities around the world. Section 4 provides insights into the different business practices that render IP law inefficient because they are allowed to frustrate its original purpose, which was to incentivize creativity through a system of fair remuneration for intellectual labour for the benefit of all. Section 5 explores the reasons that IP law is largely ineffective in providing fair remuneration, including a lack of adequate levels of coordination with other closely relevant legal regimes, such as tax, competition or international trade law. Finally, Section 6 concludes the discussion with the formulation of three principal steps to be taken to properly frame the future IP law debate in order to contribute to global peace and stability by paving the way for true incentivization of creativity based on a more sustainable, inclusive and fair remuneration of labour in the future.

2. The misconceptions in the debate about AI and IP

2.1 AI as an oxymoron and misnomer

A first, major deficiency in the AI debates is driven by mainstream media and academia, and it lies in the way the concept of AI is uncritically used and framed. AI is primarily presented as a technical tool capable of challenging and eventually superseding human intelligence now or in the near future. The debate disregards a more critical engagement with the concept, such as its qualification as an oxymoron and misnomer.⁸ It also neglects alternative concepts, such as ‘intelligence augmentation’, which places more emphasis on human intelligence.⁹ The main result of such a misconception is that the debate fails to critically reflect on not only the wider effects of AI on human intelligence but also, notably, on human lives.

This misconception leads to a second, related deficiency that lies in the dominant regulatory approach that treats AI as a technology. It also leads to the fragmentation of the debates along the lines of existing legal categories, such as IP law, competition law, labour law or tax law. The widespread separation of fields that are intrinsically linked to isolated debates reinforces existing levels of fragmentation between international organizations and their national counterparts. Internationally, these organizations do too much in isolation and not enough in common with regard to AI.¹⁰ At the national level, the same thing happens in areas of law such as notably, private law,¹¹ criminal law,¹² labour law,¹³ competition law,¹⁴ world trade law,¹⁵ health law,¹⁶ IP law¹⁷ and fundamental rights law,¹⁸ which threatens to further erode the integrity of law. It also prevents the law from carrying out its role in building the foundations of the future legal, political, economic and social order.¹⁹

Similar misconceptions guide the debate on the impact of AI on IP. In this context, an incorrect assumption is made about the proper functioning of the IP system before the rise of AI. Although GenAI was rightly called a ‘stress test for copyright law’,²⁰ the woes of copyright and IP law as a whole began long before the Dartmouth Conference was even credited with the baptism of AI, which took place in the summer of 1956.²¹ It is, therefore, naïve to assume that the future challenges for IP resulting from AI can be effectively addressed without dealing with the deficiencies inherent in the global IP regime that also troubled the past.

Among these deficiencies is the significant failure, or even the lack of will, to define ‘creativity’.²² This is the same mistake as has been made with the concept of ‘intelligence’, thereby necessitating that greater attention be paid to its diverse and multiple forms²³ Similar to intelligence, creativity is an important part of life per se and ‘the processes cannot be easily partitioned’.²⁴ This is precisely what happened in the realm of IP law, in which ‘intellectual property’ as a fruit of creativity has become fragmented. More concretely, IP law has traditionally been bifurcated into an area consisting of both commerce and industry and a second, single area of culture and the arts.²⁵ The bifurcation is largely reflected in the separation of the areas of patent and trade mark law, which are covered by the 1883 Paris Convention, from the cultural and artistic focus of copyright law, which is governed by the 1886 Berne Convention.²⁶ In this regard, the entire regime of IP law is based on an incorrect understanding of creativity that regards it as the sole privilege of the arts and various artistic sectors. A better way of conceiving of ‘creativity’ in legal terms would be to base it not on what you do, namely, whether you are a painter or singer, but rather on how you do it in the sense of ‘the [creative] quality that you bring to the activity that you are doing’.²⁷ In the same spirit, it was found that ‘any innovation – including scientific and technical innovations – of any sort in any industry is creative, and, in such terms, any industry is, therefore, potentially a “creative industry”’.²⁸

2.2 The myth of the poor but creative artist

The false dichotomy of creative versus commercial work is also manifested in the perpetuation of the age-old myth of the poor but creative artist.²⁹ This myth propagates the unfounded idea that in order to be creative, one has to be poor.³⁰ It has also obstructed the quest to institute fair remuneration for intellectual labour. By contrast, the IP law system fails to incentivize creativity with a system of fair remuneration for intellectual labour and fails to ensure a more equitable remuneration of all labour in society. The fact is that the debate about fair remuneration is not exclusively about artists, but about everyone’s fair remuneration. Similar to the overly simplistic binary separation of culture and the arts from trade and commerce,³¹ the overall focus suffers from excessive levels of specialization in legal practice and education. Excesses in specialization directly result from the neglect of the value of the generalist lawyer’s ability to oversee the broader implications of problems and formulate solutions to them in a mutually coherent way.³²

Thus, many debates about AI and IP are separately focused on AI and copyright, on AI and patents and on AI and trade marks. Such separate approaches not only fail to treat IP law as a consistent legal field but also obstruct a holistic assessment of the IP legal regime in the context of other legal regimes, such as those of competition, health or global trade law. Additionally, IP law at the international level has been aptly described as the ‘messiness’ or ‘territorial mess’ of international IP law.³³ This is further aggravated by the narrative of a race to regulate AI, which means that not only is the AI governance debate largely fragmented into national regulatory debates but so are all IP- and AI-related regulatory debates.³⁴

3. IP law and fair remuneration

3.1 Generative AI and IP law

Instead of framing the debate on the impact of AI by focusing on incentivizing creativity through fair remuneration in the wider context, the debate has been centred around three principal issues associated with GenAI and related lawsuits.³⁵ These three issues are the question of the IP protection of AI-generated works,³⁶ the liability resulting from the infringement of IP-protected works by of AI-generated works³⁷ and the infringement of existing IP-protected works based on the training of AI foundational models.³⁸ While these issues address important challenges for IP law posed by the novel possibilities brought about by generative AI, they are not a complete account. They neglect the overall challenges posed by AI not only for IP law but also for society, the economy, work and life as a whole.³⁹ It was rightly argued that it ‘would be preferable for legislators to engage in thorough discussions with stakeholders to develop a considered regulatory plan first, which does not necessarily have to revolve around copyright’.⁴⁰ Even IP law alone has more far-reaching repercussions for society as a whole than it is often assumed, which is why the debate about AI and also about the IP system as a whole needs to be approached and discussed through ‘the prism of theories of justice’.⁴¹

Within the realm of IP law, there is one central issue notoriously absent from the debate, which is the financial aspect in a generalized societal context. There is also little consideration for the future of work altogether and what role creative work will play.⁴² It has been noted that the financial function of IP law is both as an ex ante incentive to create something that is beneficial to society and as an ex post or ‘after completion’ reward for having added something of value to society, for which creators are also entitled to receive compensation for their time and money as well as for making the work publicly available.⁴³ Unfortunately, none of these three functions are adequately realized nor are they sufficiently considered in political debates and governmental efforts. It is time to seriously ponder the future role of IP law in the fair remuneration for work carried out in the economy, as has been suggested by only a few academic writings.⁴⁴ There is a sweeping lack of concrete initiatives that aim to realize the right of fair remuneration as a ‘central employment-related right and a cornerstone of the right to an adequate standard of living’ within the context of the valuation of labour in society, let alone the global economy.⁴⁵ Worse still, when provisions to this end are formulated and enacted, they are often neutralized by contractual or other unfair practices that are disproportionally unfavourable to creators.⁴⁶

In sum, the present IP system largely disincentivizes creativity and does not provide fair remuneration because it benefits big corporations rather than individual creators.⁴⁷ In its current form, it leads to greater inequality as economic disparity is built into the system.⁴⁸ As an example, the revelation that many copyright-protected works have been used for the training of AI was called the biggest heist in history.⁴⁹ History has repeatedly shown that all the hopes promulgated by techno-optimism about new technologies instantly and unconditionally bringing higher wages and greater prosperity for all have been frustrated.⁵⁰ They were destined to be frustrated at first because technological progress never automatically brings benefits. It initially enriches ‘a small group of entrepreneurs and investors, whereas most people are disempowered and benefit little,’ and it is only if social power changes that potential benefits brought by new technologies also reach the wider society.⁵¹ A more constructive debate about the impact of technology on equality levels in society is non-existent. For these reasons, the debate about IP law must return to its roots and find a new way to ensure fair remuneration not only for all but also between human and machine labour (or those manufacturing and operating the machines).

3.2 International IP law and inequalities

Reports about global wealth inequalities are staggering, and the inequality gap continues to widen. The global economy and international legal framework governing it create more inequality both between and within countries or markets. For instance, since 2020, the five richest men in the world doubled their fortunes, while almost half of the world has become poorer.⁵² These horrendous examples could be almost infinitely extended and already reveal unbelievable levels of social injustice. Contradicting a widespread belief, inequalities are not due to a person’s own responsibility but are owed ‘to factors beyond the control of individuals’.⁵³ It has been stated that inequality is not due to a natural phenomenon or that it is simply inevitable, but is rather the ‘result of law and public policy choices’.⁵⁴ In addition, IP law must be critically studied for its contribution to increasing global levels of inequality, especially since it was included in the World Trade Organization’s single package in the form of the Agreement on Trade-Related Aspects of Intellectual Property Rights.⁵⁵

Despite being an understudied problem, the current IP system has been found to be a little meritocratic in the sense that it ‘provides in some cases an unfair advantage to people who do not have anything to do with the creation of the IP’.⁵⁶ This is why IP laws must be amended to better navigate ‘the delicate balance between rewarding individual merit and ensuring broader societal equality’.⁵⁷ Various IP instruments, such as patents, copyrights and trade secrets, have adverse effects on equality due to their ‘potential to establish temporary but significant market power in specific industries and products, which seems correlated with rising market concentration, profit shares, and manager salaries’.⁵⁸ Moreover, the study of the causes of inequality among countries is masked by misleading terminology, such as the so-called ‘developed and developing countries’ dichotomy, which artificially divides the global community.⁵⁹

It is equally important to study inequalities related to IP within countries, which has so far been glaringly neglected.⁶⁰ This should also include the study of sectoral inequality among industries at the subnational level as well as inequalities within and among sectors and countries related to income, wealth and consumption, as measured, for instance, by the Gini coefficient.⁶¹ In addition to the importance of income inequality in the context of IP, it is also worthwhile to look at various industry-specific, anticompetitive practices that lead to inequalities within the same sector or even the same job.

4. The circumvention of the international IP Law system

4.1 The allures of the ‘star system’

To illustrate the negative effect of IP law on creativity and equality, the so-called ‘star system’ can be cited, which was originally developed for the nascent film industry as a tool to minimize risks and maximize profits. It was invented by the film producers and not stars themselves with the aim of creating ‘around outstanding performers such a halo of glamour that the individuals acquire a massive personal drawing power’.⁶² While this system may have effectively increased profits by drawing more viewers into the cinemas, it has evolved and has also proven to be economically beneficial in various other, seemingly paradoxical, ways.

It is important to note that, in the 1920s, the Hollywood film industry was already concentrated in the hands of the big five studios, which used stars as a ‘vital asset in maintaining the hegemony of the major studios over the whole domestic film industry’.⁶³ The star system has been widely studied, but the focus has usually been more on the social and cultural than the economic or industrial effects. In economic terms, it has been observed that stars represent ‘a form of capital possessed by the studios’, a promise ‘against loss on investment and even of profit on it’, and they were largely ‘used to sell films, to organise the market’.⁶⁴ As a system, however, there is a financial benefit of the star system that is additional to the mere attraction of larger crowds to the box offices, one which relates to the issue of inequality and fair remuneration for labour. Seen from this angle, the star system can be summed up by the motto ‘remunerate one or a few persons astronomically or far too much so as to save money by paying most persons nothing or at least not adequately’. Put differently, the reasoning goes that it is cheaper to pay a single person USD 30 million for a role in a movie than to pay a fair wage to 100 000 unknown actors aspiring to be a star.

This explains why in the film and most creative industries, creative individuals, such as actors, musicians and painters, cannot sustain their livelihood with their passion but need to work other (underpaid) jobs to sustain their passion for their creative work.⁶⁵ This undesirable situation (for most but the stars) is sustained by creating a constant oversupply based on the psychological appeal of becoming a star. In 1979, it was reported that ‘roughly 90% of Hollywood’s Screen Actors Guild membership of 23 000 earned less than a living wage and among the membership of Equity in the UK, 70 per cent of members are unemployed in any one year’.⁶⁶ Another study combined the data of millions of actors and actresses from 1888 until 2016, which confirmed a situation of endemic scarcity of resources, a rich-get-richer mechanism of job assignation and that ‘high productivity is not necessarily based on merit and is likely to be a network effect’.⁶⁷

This is clearly not a problem of the past or one affecting only a small sector of the work force in the economy. On the contrary, the economic dynamics of the star system have long ago begun to spill over to most other industries and virtually any kind of labour.⁶⁸ Today, the astronomic salaries of stars are not limited to film or pop stars, but are regularly found in other sectors, such as sports, management, academia or even politics. This means that there is a serious pay gap among persons working in the same profession and perhaps even on the same job. It is a system in which a footballer on the field earns USD 285 million in 2024 compared to an average base salary of professional soccer players estimated at USD 52 219 per year.⁶⁹ Too little is known about the discrepancies in pay received in other jobs in which salaries are subject to individual but confidential negotiations, such as governmental jobs, or in academia where the phenomenon has been called ‘academostars’.⁷⁰ Overall, this trend towards the star system has been described as follows:

As top performers become increasingly well known, team owners, motion picture companies, major publishers, boards of directors, and other big organizations compete for their services with big-money offers. This leaves less for everyone else.⁷¹

This trend has led to a situation in the organization of the labour market and society as a whole in which the ‘winner takes it all and leaves much less for the rest’.⁷² This system is supported by a combination of factors, most of which can be reduced to economic reasons. One of them is the growing trend towards greater concentration of power, including agenda-setting power and the shaping of public opinion, in the hands of fewer and fewer persons. This situation effectively prevents solutions to many problems of existing inequalities, such as the gender pay gap, or additional pay gaps caused by other factors, such as age, ethnicity or culture.⁷³ Overall, the core question about the future of IP law is to not limit the considerations to economic factors alone, but to take into account the wider societal ramifications. After all, even economists recognized that the ‘real phenomena of human economy, as paradoxical as it may sound at first, are to no small extent of an uneconomic nature’.⁷⁴ Hence, IP ought to be a system that incentivizes successful creations to the benefit of all rather than merely protect those that were successful in the past.

4.2 Economic concentration and lack of legal regime coordination

There ‘is an enormous amount of money at stake with intellectual property rules’.⁷⁵ This was illustrated by the list of the 100 richest persons in the USA, whose wealth was largely made in IP-related industries.⁷⁶ Given the existing levels of vertical and horizontal concentration in the global economy, if no adequate legal countermeasures are taken, there will soon be only five big players remaining in every major industry⁷⁷ What has happened in the field of publishing, accounting, car manufacturing, the media sector and so on has also been witnessed in the sector of technology (big tech) and will soon emerge in and be further exacerbated by the field of AI (big AI).⁷⁸

This global phenomenon of market concentration of big tech, and soon big AI, is a combined failure that is not only attributable to IP law. It is largely due to a lack of systemic coordination among different legal regimes or the absence of legal regimes. The lack of global competition rules has likely allowed many of these transnational companies to become too big too quickly and has failed to deter them from excessively abusing their dominant positions.⁷⁹ This can be exemplified by the anticompetitive effects of horizontal shareholding, which means that ‘a common set of investors owns significant shares in corporations that are horizontal competitors in a product market’.⁸⁰ Such institutional investors usually have minority shareholdings in many firms in the same market and therefore evade qualification as a dominant player. Yet, their overall influence still allows them to maximize their profits by increasing the profitability of all firms in that market by softening competition among them.⁸¹ This has the same effects as forming a cartel or behaving collectively like a monopolist as it increases the total profit in the market to the detriment of the consumer.

This danger was raised as early as 1984 when it was found that ‘firms, acting in the interest of their shareholders, tend to act collusively when their shareholders have diversified portfolios’.⁸² However, nothing was undertaken by competition authorities to intervene. The same phenomenon is addressed by common ownership theory, which is concerned with the problems caused by ‘large financial institutions holding widely diversified portfolios of shares in competing companies within a particular economic sector’.⁸³ Thus, institutional shareholding is another considerable factor in widening the gap of inequalities. Finally, it is a serious problem when the industries that create the problems are entrusted with the formulation of the solutions via self-regulation. It is the same when big tech is able to dictate the discourse and terminology of legislators or ‘the debates around AI regulation are conducted as a dialogue among a narrow set of technocratic elites’.⁸⁴

Given the interconnectedness of the global economy, other legal regimes should be critically analysed as to their individual and collective effects on the increasing levels of inequality. To mention one more, international trade law plays a potentially pivotal role in the increasing inequalities, especially when it is not adequately coordinated with other areas that are often falsely designated as ‘non-trade concerns’.⁸⁵ Paradoxically, increasing inequality within countries not only raises the probability of trade conflicts among countries⁸⁶ but also affects the global economy as wars bear not only a heavy toll on human lives but also bear huge economic costs, which in an integrated global economy are also imposed on countries other than those involved in the war.⁸⁷

Other legal regimes, such as tax law, also play an important role and need to be adequately coordinated.⁸⁸ Tax law embodies the difficulty of a single area of law to effectively address a complex issue, like the one of inequality, as every area of law is intrinsically linked to the legal system as a whole.⁸⁹ Tax law also reveals the problem of a lack of global harmonization because the inequality among countries is aggravated by tax competition, which ‘undermines the integrity and distributive principles of domestic tax systems’.⁹⁰ Ultimately, inequality has multiple causes, and they are deeply rooted in more than just single laws, but include the need for effective mutual coordination of laws within any domestic legal system. Domestic legal systems must also be better coordinated to avoid global competition among them.⁹¹ To better understand the need for such coordination, particularly in view of the additional problems for social injustice and inequality brought about by AI,⁹² it is useful to cast light on the complexities of some of the existing business practices exemplifying the failures of the existing global IP regime.

5. IP law-related failures exemplified

5.1 New business models and legal loopholes

‘The plenty of my intellectual property makes me poor’ could be a modern interpretation of Ovid’s Metamorphoses.⁹³ While it may not be necessary for IP law to make you rich, it should provide some fair remuneration and certainly not make you poor. A vast body of empirical work has shown that various IP rights do not offer ‘decisive financial reward to most artists and other primary creators’ as it is ‘simply not true for the vast majority of practising professional artists, though a small proportion of star artists do earn well from royalties and other copyright earnings’.⁹⁴ What, then, are the factors that cause these frequent failures? First, many of the business models have altered the game by creating confusion through blurring the lines of distinctions. Linguistically, this practice is reflected in the rise of the use of the rhetorical figures of speech of paradoxes and oxymora.⁹⁵ It is no coincidence that numerous related concepts such as ‘culture industry’, ‘virtual reality’, ‘synthetic biology’ or ‘artificial intelligence’ are but a few examples. Even the term ‘intellectual property’ itself was qualified as an oxymoron.⁹⁶ This qualification was used to point out that ‘[C]opyright as it now stands has outlived its original purpose, and is no longer clearly beneficial to society as a whole’.⁹⁷

Multiple failures of IP law hide behind this trend, and they materialized long before the rise of AI. These failures stem from the practices of new business models to exploit the loopholes in today’s laws and transcend former, well-established conceptual boundaries of a mostly dualistic mode of thinking. For instance, the business models of social media and digital platforms have been found to be multisided and to cater ‘to demand from different sides, be that the demand from consumers or the demand from other businesses’.⁹⁸ This has effectively altered the traditional relationship between the producer and the consumer, which was captured by the concept of the ‘prosumer’, an oxymoronic portmanteau word suggesting a simultaneous process of production and consumption.⁹⁹ However, the interpretation of the prosumer as ‘the consumer who also designs and produces what s/he consumes’ is not entirely accurate when it comes to the business models applied by digital platforms¹⁰⁰ In that context, the consumer freely produces content for the platform, which other consumers can freely access, but the profits generated from this are taken by the owner of the platform, largely through revenues generated from advertising. But it is false to assume that this is the only source of revenue, as prosumers—lured by the mirage of ‘free products’—pay a high price in the form of their time and their data, with the latter having been called the ‘new oil’.

Data also have more indirect benefits, such as enhanced possibilities to manipulate consumers for commercial or political purposes, including various subliminal methods, for example, dark patterns, that are outside a person’s awareness.¹⁰¹ These manipulative possibilities open the opportunity for a further increase in profits, as they offer new venues for the exploitation of the vulnerabilities of individuals with the help of psychology and neurotechnology.¹⁰² They allow products or websites to be designed that intentionally exploit vulnerabilities developed by the evolution of the human brain.¹⁰³ It has already been found that, when it comes to consumption, humans more often act irrationally. Manipulation and marketing may explain ‘why we often pay too much when we pay nothing’ or, more precisely, ‘believe to pay nothing’.¹⁰⁴ The same finding was aptly summarized by the Internet-age and digital economy dictum of ‘if the product is free, you are the product’.¹⁰⁵ Finally, the changing environment of the producer and consumer has also been expressed by the profit paradox concerning labour. This paradox, which is also relevant for AI, states that technological change has created enormous potential for economic and social progress but has instead led to a situation in which the market power and dominance of the thriving firms are ‘not beneficial for workers’.¹⁰⁶

5.2 The creative industries and destructive exploitation

Workers do not connote labour for a specific sector. The term also applies to intellectual work, such as the academic publishing industry. Even before the advent of GenAI, profits in this industry were and still are far from being fairly distributed between the creator of an intellectual work and the distributor. Academic publishing agreements for books usually foresee a 5–10 per cent profit margin for the author and 90–95 per cent profit for the publisher, even though objectively, the time invested and work performed by the author would suggest the exact opposite allocation of profits. At the least, it should be an equal (50–50) split between the parties to the agreement. It gets worse with the publication of journal articles, for which no fees are usually paid at all, even though a single purchase of an author’s article is commonly priced at USD 30–50, the same price as a book. If an article is published open access, the fee the author pays for their own work to be published is somewhere between USD 3000 and 4000, which seems closer to the real value of the work or at least what it is worth to the publisher. By contrast, it can cost more than USD 200 to ask a journal for permission to cite one sentence from an article as an epigraph (not covered by an applicable exception).

The system gets even more exploitative given that universities pay large amounts in yearly subscription fees to the digital databases containing the publications, which scholars (including the university’s staff) provided practically for free. Despite frequent lamentations, the huge profit margins of the academic publishing industry have been said to exceed those of Apple, Amazon or Google, companies that are not precisely known for their generosity.¹⁰⁷ The situation gets even worse when the greater context of the entire business model is considered and not merely the contractual terms in the publishing agreements. Publishers do not reward the same authors for articles they write and the peer reviews they conduct for them. Even the earlier custom of gifting one or two printed copies of the volume in which an article was published has been replaced by a ‘preferential’ rate charged to the author at a 30 per cent discount.

The short study of this common practice in the academic publishing industry illustrates well the need to consider the whole context rather than merely discuss IP law. It is the business model and the societal context, such as the remuneration for academics, that should be considered in their entirety. The practice of zero pay for peer-reviewed articles, for instance, is part of the problem and potentially pertains to the realm of competition or antitrust law. This can be seen in a recent antitrust class action in the USA (finally) brought by a scholar ‘to challenge collusion among the world’s six largest for-profit publishers of peer-reviewed scholarly journals’ for allegedly having formed ‘a cartel to fix the price of peer review labor at zero’.¹⁰⁸ It is also a question of the entire educational system and labour law. For instance, while it may have been justified to pay scholars little or nothing for their academic writing if they were employed by a university, the present system is no longer in balance, considering reports about dwindling academic salaries or junior scholars having serious problems even securing a stable position in a university yet being forced to publish prolifically.¹⁰⁹ It has been observed that ‘the kinds of restructurings of academic work that are underway today not only have increased employment insecurity in academia but also may actually be producing unemployment both within and outside academia’.¹¹⁰ Finally, in academia, too, the star system works, and it is not surprising then that the few ‘academostars’ who are paid a solid salary from a tenured position at a university, plus potentially earning fees from aiding big tech companies in high-profile lawsuits, may care less about fair remuneration for academic writing and even advocate a fair use exception for the training of large language models with copyright-protected material.¹¹¹

This scenario is not limited to the academic world but is symptomatic of the lack of balance in the entire economy, both national and global. It is a reality for all types of creative artists, from painters to filmmakers or composers, all of whom exemplify ‘precarious work’ that requires ‘multiple jobs to subsidize their creative work’.¹¹²

To take the example of the fashion industry, the ‘piracy paradox’ was used to describe the failure of the copyright system as it showed that it is not the legal protection offered by IP rights that makes the difference in the profitability of a sector, but rather the business model developed around (the intention of circumventing) the rights.¹¹³ Later, the scope of the paradox was expanded to the entire economy by the notion of the ‘knock-off economy’.¹¹⁴ In general, the situation in the labour market of the arts was described by the following trends:

[L]ong-term employment has been replaced by a project-based system of production relying on short-term hiring, large parts of business risk are transferred downwards onto the workforce, and artists learn to manage risk and to stay alive through multiple job-holding, occupational versatility, diversification of job portfolios and occasional income transfers from social security or other sources.¹¹⁵

Contrary to the common perception, this short description is not limited to the arts but is applicable to all types of jobs. It confirms that many of these failures are not caused by IP law alone but must be seen in the wider context of bigger changes affecting other areas of law.

In international trade law, it is the age-old separation of public from private and international from domestic law as well as insufficient levels of international harmonization that play a harmful role.¹¹⁶ This situation of the legal fragmentation of the global market allows multinational companies to evade stricter rules at both the domestic and international levels.¹¹⁷ It also allows the segmentation of consumers along their respective purchasing power in order to maximize profits.¹¹⁸ The situation is further affected by the absence of global competition rules. As early as in a case decided by the US Supreme Court in 1948, it has been known that multinational companies have deployed various anticompetitive practices (eg block-booking, zoning, prototypes and the star system).¹¹⁹ These practices have evolved since then, as can be seen from the debate about trade in counterfeit products. This debate fails to acknowledge various aspects, such as the existence of so-called ‘real fakes’; that is, products that have entered the market through dubious ways, such as ‘ghost’ or ‘third shifts’.¹²⁰ Such real fakes help to generate additional profit, even matching or exceeding the profit of the high-end luxury segment.¹²¹ They are aided by loopholes or inadequate levels of harmonization of the issue of parallel imports.¹²² Counterfeits have even been found to create several indirect but positive side effects, such as guilt that leads to the purchase of originals at a later stage.¹²³ These factors and considerations about the role of IP laws in inequality are usually not included in the debate about counterfeits or piracy.¹²⁴

Instead of a constructive and comprehensive debate, high-profile lawsuits maintain the strict dichotomy between real and fake products or originals and their (allegedly unauthorized) copies. This helps to maintain the chimera of their difference so as to maintain the stark difference in prices. One such high-profile ‘mock trial’ was brought by Luis Vuitton against Warner Bros for an alleged trademark infringement in a motion picture based on a dialogue between two characters involving a knock-off item similar to the plaintiff’s products.¹²⁵ Another similar trial concerned the alleged plagiarism by Dan Brown in his book The Da Vinci Code, which was held just a month before the release of the movie.¹²⁶ Further evidence of its nature as a mock trial is that the plaintiffs’ book upon which the infringement was alleged was published by the same publishing house.¹²⁷ Finally, it is also curious that a news report published by The Guardian after the release of the verdict, in which the judge had also hidden a secret code, stated that ‘sales of both books rose during the trial, which has spawned headlines around the world’ and also mentioned that a film version of the novel was due to be released on 19 May 2006 starring Tom Hanks and Audrey Tautou.¹²⁸ It is amusing to note that the online report concluded with two links enabling the reader to purchase both books, that of the plaintiff and that of the respondent.

Cutting a complex story short, these few examples illustrate several IP law-related failures that show that the creative industries were able to exploit the failures to their own profitable advantage. It is hoped that the examples suffice to support the need for a more inclusive and far-reaching debate about the regulation of AI instead of the mere fragmentation of the issue into many separate AI debates, such as the debate pondering the impact of generative AI on IP law. It is already a good sign that numerous commentators have begun to critically assess the role of IP in increasing the levels of inequality. However, the process cannot stop there and must continue to seek an expansive debate among law experts in the largest possible number of legal fields. Or better still, the process must endeavour to tear down the traditional boundaries drawn between legal categories, which have been rendered obsolete by various new technologies, ranging from digitization to AI.

6. Conclusion

Cui bono? – ‘The illustrious Lucias Cassius, whom the Roman people considered the wisest and most conscientious of judges, was in the habit of asking repeatedly in trials, “who had profited by It?”’¹²⁹

The Latin expression ‘cui bono’ (to whom the benefit) ought to be more than a maxim for an investigation into the alleged perpetrators of crimes; alternatively, it should be applied to the quest for an adequate regulatory response to changes in lives caused by disruptive technologies. It has been shown that the current regulatory debates about AI are in many ways misconstrued, not only because of their uncritical use of the concept itself but also due to their fragmented focus on separate aspects and legal areas related to AI rather than on a holistic assessment of its overall impact on human lives. Even the narrower debate about the impact of AI on IP law is not framed correctly, as it is not AI that threatened to turn IP law on its head; the latter had already failed prior to the advent of GenAI to live up to its original task.

One reason for the failure of IP lies in not only the development of new, innovative but also anticompetitive business practices pioneered by the creative industries. These practices were further developed with the introduction of digital technologies and led to a situation in which IP law amounts to a system of legal expropriation described as follows:

If attorneys still defend certain creators in the name of the right to intellectual property, most of them have become specialized in the matter and counsel investors in the so-called creative industries in profiting of their extension of the aura attached to this revolutionary right. This extension is to the detriment of the principal contributors to these industries, summoned occasionally and working for free.¹³⁰

In all fairness, IP law alone is not to blame for this great injustice or economic dysfunctionality as no single area of law is capable of providing effective responses to the broader societal, political and economic challenges that have arisen from the latest disruptive technologies. It is the continuing and even accelerating trend of the convergence of technologies, products and industries that has overburdened the world’s many legal systems, which were built on the past separation of law into various subfields. As a result, legal systems are no longer able to guarantee a fair remuneration for everyone’s work in the economy, thus contributing to dangerously increasing levels of inequality, not only between and within countries but also within the same job. Consequently, it also means that the current challenges posed by AI and generative AI as well as the digital economy as a whole cannot be tackled within the field of IP law or, for that matter, any single legal field. Even within IP law, the central focus should not be on some technical details regarding the status of IP-protected works used or created by generative AI. Instead, the debate should return to the origins of IP law and realize the centrality of the human creator of a work, for not only creative work but also labour of any type.¹³¹ Thus, true to its pioneering role in innovation and progress, IP law should make a constructive contribution to a debate about the global regulation of AI by formulating new ideas and elaborating relevant concepts that allow the establishment of a global system for the fair remuneration of all types of labour in the future. A first step in this direction is a renewed understanding of creativity, one which does not depend on what you do but how you do it, which can provide a new basis for fair remuneration in a future society in which—because of AI and robots—there will be no more work as all humans will be creators.

For governments and regulators aiming to reform their IP laws to tackle the challenges brought about by AI, it is therefore advised to draft amendments to existing IP laws in close consideration of other areas of law and critically assessing their impact on the legal system as a whole. To this end, they should engage in a broad multi-stakeholder dialogue and also set up designated working groups or involve or set up competent law reform commissions to balance regulatory with deregulatory efforts. Most of all, it is necessary to draw up a future-proof plan and vision for the future designation, role and fair remuneration of work and then consistently pursue it by legal means.

Acknowledgement

The author thanks Tianxiang He and Yang Chen for convening the workshop ‘Comparative Perspectives on AI and Copyright Law: Evaluating Hong Kong’s Policy Responses in the AI Era’. I am also grateful to Célia Filipa Ferreira Matias, Han Tianyu, Wang Runhua and Tianxiang He for the useful comments received on an earlier draft of this paper. Last, the author acknowledges the financial support provided by the University of Macau for a research project entitled ‘Law, Language and Technology: On the Future of Human Cognition’ [MYRG-GRG2023-00046-FLL].

Footnotes