The Caribbean  Journal of Law , Policy and Social Change  was started in 2010  as  the Caribbean Law Journal Online to promote writing and scholarship in law and policy the Caribbean community. In recognition of  the lasting and transformative value of social change , and the need to incorporate policy practices that create positive impacts in legislation, policymaking and implementation, the journal has been rebranded as the Caribbean Journal  of Law, Policy and Social Change. This is a peer reviewed journal sponsored by the Caribbean and Americas Intellectual Property Organization (www.caaipo.org). We invite you to contribute your expertise in collaboration with those who support the concept of scholarship for development of the Caribbean and its peoples. We welcome submissions on interdisciplinary research which examines current and future areas of many field in the humanities including economic, environmental , sociological , climte related, gender isuues, food security , intellectual property and others. The aim is to broaden the discourse on the future of the region and to help move individuals, groups and scoieties in the region towards a more just and sustainable future.

We welcome your submissions and look forward to adding your contributions to the discourse.

Intellectual Property, Technology and Innovation for economic development and Social Change in the Caribbean Community (Caricom).

By Dr. Abiola Inniss Ph.D. LLM

Abstract

Caribbean countries, particularly those within the formal structure of CARICOM (The Caribbean Community), find themselves at the juncture of an urgent need for greater self sufficiency for food security, mitigating the effects of climate change, and improving the quality of life for citizens.  In an age of advanced and continuously advancing internet, artificial and mechanical technologies, the region is severely lacking in resources that can provide significant benefits in bolstering economic growth, protecting and providing new livelihoods, and reducing the challenges of food insecurity. Many of these technologies are already in use in other parts of the world and the majority of this knowledge is protected by intellectual Property Rights (IPRs).

The concept of intellectual property (IP) has been linked to the creation of change in communities through enhanced economic opportunities and tangible benefits where it can be properly implemented and utilized. Some opponents of bolstered intellectual property rights argue that their restrictive nature prevents individuals from accessing and using resources freely, thereby preventing them from reaping economic benefit and furthering the development of technologies which are useful to them.  This essay posits that In order to successfully exploit intellectual property within the context of maximizing economic growth in the Caribbean community, it is important to invest in areas that are IP and knowledge based intensive. This in turn can foster the production of goods and services that are rich in innovation and make meaningful changes for consumers, businesses, and the larger community across the Caribbean region.

Social change and Intellectual Property.

            The concept of social change has been described by some thinkers as comprising thoughts and actions which move progressively from one epoch to another and sometimes bring improvement to the community under consideration. It is also notable that as the frequency and intensity of dramatic social change continues to increase in today’s world It is also thought that social change can be regressive and encompass negative developments for a community[1]. The quality of change and the direction that it takes must therefore be defined within the context of this essay, since it should not be assumed that all social change is positive. The idea of social change can be associated with positive impact on the community or on the society at large, however the reality of change of any kind is that it brings a different state of existence into being which can be positive or negative. The ideas of social change have historically been represented in three major cycles in scholarly teaching in different societies. First there is the idea of decline and degeneration, then there is cyclical change in which there are recurrent phases of growth and decline, and last, there is the idea of continuous progress. In any society all three of these patterns can be observed in relation to a number of aspects of its existence so that it is important to delineate clearly amongst them[2].

            The construct of intellectual property may also be linked to the creation of change in communities through enhanced economic opportunities and tangible benefits where it can be properly implemented and utilized. Some opponents of bolstered intellectual property rights argue that their restrictive nature prevents individuals from accessing and using resources freely, thereby preventing them from reaping economic benefit and furthering the development of technologies which are useful to them[3]However, there is little evidence of any significant technological developments in the absence of IPRs in the Caribbean region. In fact, the Caribbean region has lagged behind in technological developments in all of its member countries whether or not these countries have current IP laws. In investigating the levels of innovation in the four largest economies in the region, namely, Guyana, Jamaica, Trinidad and Barbados in a 2017 study, Guyana was found to have the greatest amount of innovation while its IP system was the weakest in the Caricom region[4]. There is little current evidence to suggest that this has changed for Guyana and the WIPO Global Innovation Index 2024 does not provide any rankings for the country. Over the past 30 years communities across the Caribbean have been plagued by a decline in export diversification and a growing specialization in the markets to which they export. These exports are mainly comprised of low value-added goods, raw materials and a few commodities. Tourism, goods and services also receive stiff competition from other parts of the world and are affected by global financial fortunes[5]. Additionally, the small population size and the lack of emphasis on science and technology development by regional governments have made advancements in this area miniscule in comparison to the use of imported technologies by the population.

       There continues to be some considerable debate by international scholars as to whether the development of innovation and technology in the global south (developing countries) needs to be prompted by government or there can be dependable development from individuals and small to medium sized businesses which will propel it in a manner that will change the technology landscape in major and meaningful ways[6]. Questions also arise as to whether developing countries can use IP to advance their development strategies relevant to sustainable development and to meeting international goals such as those set out by the United Nations on poverty and hunger[7].It is important to recognize that sensible, robust public policy and administration are key elements in addressing these challenges, and in the case of Caricom can be a critical means of bolstering competitiveness using IP, innovation and technology.

 Economic analyses by international scholars carried out over the past twenty years have led to different conclusions about the effects of IPRs on the economies of developing countries. In some instances, the findings were that countries with greater levels of FDIs and stronger IP protections showed stronger growth in technological developments and foreign countries were more likely to invest there by opening factories and outsourcing manufacturing. Conversely countries which engaged in imitation of technologies had a more difficult time attracting investment and foreign capital which were needed to enhance their economic development[8].

            While there have been suggestions that within the context of the Caribbean countries that technological development can be best advanced by imitation followed by innovation, this implies that government policies would have to incorporate the unlicensed use of intellectual property often in contravention of national laws or/ international agreements. It is important to note that all Caribbean countries are signatories to the WTO TRIPS agreement, and that many of them have enacted at least the minimum requirements into their IP legislation. Additionally, some of them are signatories to a number of IP treaties which require certain minimum standards and reciprocal arrangements for protections. The international legal considerations and basic diplomatic concerns of unlicensed imitations are therefore of critical importance.

The Diplomacy of Intellectual Property

           International Intellectual Property law and Policy and its attendant diplomatic protocols and concerns can be considered a very important part of international relations. The simple explanation is that IP is an essential tool of economic growth in developed countries and has been so recognized as early as the 1623 British Statute of Monopolies which was later replaced by the Statute of Anne, 1710[9]. This central tenet is also enshrined in the United States Constitution Article 1 Section 8, Clause 8[10]on intellectual property as follows “To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” In the Caricom (Caribbean Community) region, the 1987 Constitution of Haiti [11]at Article 38, explicitly stipulates that artistic, literary or scientific works are protected by law, and may be the only Caricom state in which this protection is embedded in the constitution.  The seriousness with which this branch of regulation has been treated by some policymakers is a clear acknowledgement of the importance of controlling the amount, extent of access and, usage of intangibles related to creativity and technological development. It is also a major aspect of trade negotiations and over several decades has been hotly debated as to whether its exploitation benefits developed nations more than developing ones.

            The TRIPS Agreement administered by the World Trade Organization[12] has been a hotbed of argument on both sides of this contention, and there is still some resistance to writing its minimal requirements into the domestic laws in some developing countries. In some cases, countries have acquiesced to the TRIPs agreement as members of the World Trade Organization (WTO) but refuse to enact legislation or update laws which reflect the requirements of the TRIPs agreement. There is in effect a signing on to the agreement but not a signing in to it.  Some developing countries have adopted just such a stance and continue to avoid the issue of strengthening intellectual property rights legislation and enforcement whilst managing to engage in trade and other activities. The issue remains whether these countries are still able to advance development at a pace that benefits them as a whole instead of the promulgation of a narrow agenda based on limited views of IP rights resulting in a balance of power that is detrimental to the country’s economy. It is in fact the balancing of these very interests at the local and international levels that has proved daunting to some policymakers.

Some specific difficulties faced by developing countries in international IP negotiations.

Developing countries have concerns about intellectual property (IP) negotiations for a number of reasons, including: 

(a)    Economic impact:

Developing countries are concerned that stronger IP protection would

• • Drive up prices 

• • Allow foreign interests to benefit from indigenous knowledge and biological resources 

• • Limit growth in developing nations 

• • Be an extension of monopolistic practices by multinational companies. 

(b)    Access to medicines:

Developing countries are concerned that stronger IP protection could impede access to medicines. This debate became particularly fierce during the COVID19 pandemic 2020.   

(c) Enforcement standards:

Developing countries often negotiate agreements with high levels of IP protection and enforcement standards, but these agreements often lack concrete provisions on licensing and cooperation.

(d)   Political and Economic tools:

Some argue that developed countries use IP as a political and economic tool to prevent others from using it. 

Other issues in IP negotiations include: 

• The complex relationships between Intellectual Property (IP), Genetic Resources (GRs), Traditional Knowledge (TK), and Traditional Cultural Expressions (TCEs). 

• The need for negotiators to have a high degree of substantive expertise.

• The need for extensive coordination and policy coherence at the national level.

• The need to protect geographical indications to avoid misleading the public, and to prevent unfair competition. 

As outlined above, the diplomacy of intellectual property is fraught with issues of politics and economics at individual country and regional levels. In the case of Caricom, The Revised Treaty of Chaguaramas[13] which is the legal instrument establishing the Caribbean Community (Caricom), provides in its Article 66 for the establishment of a regional administration for intellectual property rights with the exception of copyright, but governments have been unable to agree on the creation of such a mechanism. It is clear that conflicting policy interests and lack of political will have stymied the process to the extent that the only regional activity on intellectual property over the past decade at minimum has come from the European Union and other bodies which have created widescale programs to promote the knowledge and use of intellectual property rights at country levels in collaboration with Caricom. The difficulty with this is that the agenda of these international bodies does not take into account local idiosyncrasies and while the general policy is to encourage the use of local provisions for IP, it is not necessarily concerned with whether the available resources are adequate or function in a way that makes it easily accessible or makes sense financially for those seeking the protections. Indeed, the latter is solely in the purview of the local policymakers who are often less than enthusiastic about addressing these issues for reasons previously delineated above.

Creating a strategy for social change through intellectual property for the Caribbean Community.

The basis of social change through intellectual property is that the positive aspects of implementing it as a tool for benefit to communities and individuals must be highlighted and implemented. There may be serious misgivings by policy makers as to the economic impact of IP implementation with regard to international economic relations, such as foreign interests benefitting from traditional knowledge, or onerous impacts on local prices for goods and services, and the potential retardation of development of local industries, and restricted access to medicines. Conversely, there are significant benefits to be had from the implementation of intellectual property rights.

Intellectual property (IP) can help developing countries, especially in the Caricom region, to advance their development strategy in several ways, including: 

• Promoting innovation

IP protection allows innovators and investors to recoup their investment in bringing new products to market. This encourages domestic innovation and foreign direct investment. 

• Supporting creative industries

IP tools and resources help creators retain control over their work, secure fair revenues, and incentivize further creativity. Creative industries like music, film, software, design, and publishing can generate job opportunities and foster local innovation. 

• Attracting investment

Strong IP frameworks increase investor certainty and lead to increased funding for local creative projects and industries. 

• Diversifying economies

IP can help developing countries diversify their economies and exports. 

• Participating in the global economy

IP systems can create a framework for developing countries to participate in the economic activities in the global economy. This may be intra-regional or at a greater global level amongst other developing or developed countries.

It is within this context that Caricom countries should look for opportunities to use intellectual property to benefit businesses, communities, and their economies as a whole. In creating a strategy for IP the policymakers must first establish a vision for at least a medium term development plan which includes a focus on science technology and innovation. This may seem to be daunting given the availability of resources. However, if one considers the example of neighboring Cuba which in spite of its economic woes still dedicates a sizeable amount of its GDP to science was approximately  0.32 % of its GDP as at 2021.This is interestingly, higher than that spent by Kuwait which showed 0.08 % at 2021,and in similar vein to China Macao special Administrative region which showed 0.38 % at 2021[14] (UNESCO, 2022).The dedication to science and development of the Cuban government in consideration of the economic circumstances, still provides a clear example of an adaptable approach that can be considered and even implemented in a sustainable manner.

Most Caricom countries do not have a percentage of GDP allocated to science and technology with the exception of Trinidad & Tobago which registered 0.05 % at 2021[15].The figure for Latin America and the Caribbean on average was 0.55 %. This is a clear indication that research and innovation are not necessarily policy priorities in the Caricom region and therefore the issues of IP and the development of technology have not been examined and approached with advancement and competitiveness in mind. For most of the countries in the region that have enacted legislation, it has been the result of the need to comply with elements of the TRIPs agreement in fulfillment of WTO requirements for trade relations especially as relates to Foreign Direct Investments (FDIs) in imports and exports, tourism, and hospitality.

Innovation occurs when people have the resources and the willingness to create new things. The necessary resources often vary greatly from one industry to another, and policy measures also have an impact on this. The National Innovation Pathway 2023 of the United States of America[16] is a White House office of Science and Technology Policy paper which was created in collaboration with the US department of Energy and the US department of State demonstrates an approach that facilitates an equitable distribution and availability of supporting resources to all innovators and entrepreneurs. This is obviously tied in with the economic plan which fosters innovation at all levels, encouraging entrepreneurship across the economic spectrum, and not simply focused on large industrial concerns.

Elements of a national strategy.

A national strategy for intellectual property and innovation should be crafted to encompass policies for the development of innovation and technology alongside programs that promote the development of entrepreneurship and creativity within communities in the country. Universities and other technical institutions need to be prioritized for funding and other resources such as technical collaborations with other international institutions and companies. As in the case of the developed countries discussed above, a national strategy for economic development, innovation and technology must incorporate appropriate policy and legal frameworks which also allow for more equitable distribution of resources to encourage endogenous creativity and problem solving. This will help to decrease the gap between those who have access to training and funding and those with less resources. It will also help to generate solutions to local problems.

Developing a regional intellectual property and innovation strategy, and the new UN Global Digital Compact

Advocacy for Intellectual Property Rights does not translate into blind transport unto the innovation train. The fact is that IPRs themselves are not sufficient to encourage innovation. As noted before there is need for policy and funding that targets innovation and which are protected and bolstered by measures that give entrepreneurs and innovators broad support. Intellectual property is but one aspect of a comprehensive plan for the development of innovation and technology which will impact the lives of communities and the national economies of developing countries in a manner that is uplifting and beneficial.

 In order to successfully exploit intellectual property within the context of maximizing economic growth, it is important to invest in areas that are IP and knowledge based intensive. This in turn can foster the production of goods and services that are rich in innovation and make meaningful change for consumers, businesses, and the larger community. If this situation is to be achieved, it is imperative to develop a comprehensive knowledge optimization strategy for each country within the Caricom region.

 At this point it is important to include the TRIPS agreement (Trade Related Aspects of Intellectual Property Rights) administered by the World Trade Organization ( WTO) [17]which sets out the minimal requirements for IPRs for countries that are members of the WTO. All Caricom countries are WTO members, and this membership is a critical component in their ability to do business internationally. WTO members cannot legislate below the minimal standards of TRIPS without incurring the risk of dispute settlement under the Dispute settlement Agreement which would have serious consequences for the countries involved. The TRIPS agreement contains some flexibilities on how the rules may be implemented and in what circumstances these allowances can  be applied. For example, under these flexibilities WTO members may exploit creative means of transposing into national law and practice those concepts which the TRIPS agreement enunciates but does not define. These may include concepts of novelty and inventiveness or situations of emergency such as those related to public health or biodiversity protections.

As a result, the TRIPs agreement, to the extent that it has been implemented by the countries, affects the ways in which IPRs are viewed and utilized. Across the Caricom region there are varied levels of implementation of the TRIPS requirements which have resulted in some instances in the enactment of full suites of IP legislation across the range of IPRs and in others (e.g. Guyana) the laws remain the same as at pre-independence 1966.

The United Nations Global Digital Compact[18] is a new initiative formalized on December 24,2024 which aims to establish principles, objectives, and actions for an open, free and secure digital future for all peoples. According to the policy brief it is based on universal human rights and intends to advance the attainment of the Sustainable Development Goals. This new initiative is based on the understanding that the digital divide between developed and developing countries has increased greatly with the advent of new artificial intelligences which have created a much wider chasm in economic benefits from innovation and technology in all areas in which they are deployed. These include of course medicine and medical technologies, business and enterprise, manufacturing, agriculture and other sciences which are critical for developing countries. It recognizes that digital technologies have moved beyond internet and mobile technologies to generative artificial intelligence, autonomous artificial intelligence, blockchain systems, digital currencies and quantum technologies. The UN has established a new Office for Digital and Emerging Technologies as of January 1st, 2025, to oversee the goals of this new initiative. At this time Caricom has not pronounced any new initiatives which will examine this fast-paced technological growth and make any recommendations for policy and development concerning it.

Conclusion.

The importance of innovation and technological development in the developing world is a topic of significant concern, especially in the face of serious challenges in to the survival of communities from climate change, food insecurity, communicable diseases and non- communicable diseases such as HIV and AIDS and other maladies. Modern technologies often help to alleviate and even eradicate some threats to the very existence of communities. It is critical that appropriate and relevant resources become available where they are needed, from medicines and treatment options and technological devices to the cultivation of food crops management, and farming techniques that support communities in providing food security. Many of these technologies are available in the developed world and some have been deployed in developing nations through international Aid agencies with local adaptations to make them relevant and successful in the communities that employ them. One example is the food farming practices undertaken by the Kenyan government, known as Kenya Climate Smart Agriculture Project (KCSAP)[19].

 The KCSAP program utilizes available technologies from water management, selection of crop varieties that are climate resilient, drones for precision farming and apps that give farmers access to information and technical assistance. As a result, intellectual property plays a key role in the further development of applicable and useful technologies since patents, industrial designs and other applicable rights help innovators to retain ownership and to distribute rights in a fair and equitable manner.

Caricom countries and the regional body itself need to rethink the way they strategize for the future of the region and its communities. The Kenyan model mentioned above is a key example of how innovation and technology can revolutionize food security and economic prosperity in a developing country. It is a model that can be adapted for regional development, and which should be considered by regional and national authorities. Additionally, as mentioned above in discussing the UN Digital Compact the relevant  new digital technological developments need to be harnessed and deployed in the region  expeditiously and as  far as possible.

In spite of the challenges faced by the nations in Caricom, there is evidence that by reshaping thinking and policy approaches to strategically implement technologies that can improve food security, mitigate climate change and improve the distribution and availability of medicines and health care, the region can attain a higher standard of living for its citizens. Collaboration with extra-regional scientific communities, along with endogenous innovations are the key components to such a strategy. Intellectual Property law and policy help provide a framework for encouraging adequate compensation to innovators and protections for creating the much-needed technology.

References

Cambridge University Press. (2017). The Cambridge Handbook of Sociology : Core aeas in sociology and the development of the discipline. Cambridge, UK: Cambridge University Press.

Chen, Y., & Puttitanum, T. (2005). Intellectual property rights and innovation in developing countries. Journal of development Economics, 474-493.

Ezell, S., & Cory, N. (2019, April 25). The way forward for intellectual property internationally. Retrieved from Information Technology & Information Foundation: https://itif.org/publications/2019/04/25/way-forward-intellectual-property-internationally/

Gervais, D. (2009). Of Cluster and Assumptions: Innovation as part of a full TRIPS implementation. Fordham Law Review, 2353-2377.

Inniss, A. (2017). Examining Intellectual Property Rights, innovation and technology within the Caricom Single Market and Economy. United States: Walden Dissertations and Doctoral Studies 4223. Retrieved from Kenya Climate Smart Agriculture Project: https://scholarworks.waldenu.edu/dissertations/4223

Kenya Clinate Smart Agriculture Project. (2025). Retrieved from Kenya Climate Smart Agriculture Project: https://www.kcsap.go.ke/

Lorenczik, C., & Newiak, M. (2012). Imitation and innovation driven development uner imperfect intellectual proerty rights. European Economic Review, 1361 -1375.

National Innovation Pathways of the United States 2023. (n.d.). Retrieved from National Archives and Record Administration: https://bidenwhitehouse.archives.gov/wp-content/uploads/2023/04/US-National-Innovation-Pathway.pdf

Office, U. S. (2024). National Strategy for Inclusive Innovation. Retrieved from US Patent and Trademark Office: https://www.uspto.gov/sites/default/files/documents/NationalStrategy.pdf

Overview The TRIPS Agreement. (2025). Retrieved from World Trade Organization: https://www.wto.org/english/tratop_e/trips_e/intel2_e.htm

Revised Treaty of Chaguaramas. (2025). Retrieved from Revised Treaty of Chaguaramas: https://treaty.caricom.org/

The Constitution of the Republic of Haiti 1987. (2025). Retrieved from WIPO LEX: https://www.wipo.int/wipolex/en/legislation/details/7837

The Constitution of the Republic of Haiti 1987. (2025). Retrieved from WIPO LEX: https://www.wipo.int/wipolex/en/legislation/details/7837

The Statute of Anne ; April 10, 1710. (n.d.). Retrieved from Yale Law School, The Lillian Goldman Library, The Avalon Project: https://avalon.law.yale.edu/18th_century/anne_1710.asp

UNESCO. (2022). Science,technology and innovation : 9.5.1 Research and development expenditure as a proportion of GDP. New York: UNESCO. Retrieved November 4, 2024, from https://data.uis.unesco.org/index.aspx?queryid=3684

United Nations. (2025, March 11). United Nations Department of Economic and Social affairs , sustainable development goals. Retrieved from United Nations Deapartment of Economic and Social Affairs: https://sdgs.un.org/goals

United States Government. (2025). Constitution of the United States. Retrieved from United States Senate: https://www.senate.gov/about/origins-foundations/senate-and-constitution/constitution.htm

Weinstein, J. A. (2010). Social Change. Lanham, Maryland: Rowham & Littlefield.

 The Sovereign Archive and the Digital Plantation: A Critical Analysis of U.S. Fair Use Hegemony and the Extraction of Caribbean Cultural Memory.

By Dr. Abiola Inniss Ph.D. LLM

Abstract

The rapid ascendancy of Generative Artificial Intelligence (GenAI) has precipitated a global crisis of copyright, ostensibly centered on the tension between human creativity and machine learning. However, beneath this technological dispute lies a profound geopolitical asymmetry. This article argues that the extraterritorial application of United States "Fair Use" doctrine (17 U.S.C. § 107) to the training of Large Language Models (LLMs) constitutes a form of "digital colonialism" that disproportionately impacts the cultural sovereignty of the Caribbean Community (CARICOM). By treating the digitized cultural heritage of the Global South as "raw data" for "transformative" computation, U.S. legal frameworks re-inscribe historical patterns of extraction. This paper posits that the Caribbean archive—comprising oral histories, indigenous linguistics, and communal artistic expressions like Mas—is being rendered into a "digital terra nullius," mined by foreign entities to capitalize on cultural value while bypassing the reciprocal obligations of local "Fair Dealing" statutes. I conclude that without the implementation of sovereign data licensing and Text and Data Mining (TDM) reforms, the Caribbean risks becoming a permanent periphery in the cognitive economy of the 21st century.

Introduction: The New Terra Nullius

In 1493, the doctrine of terra nullius (nobody’s land) provided the legal fiction necessary for European powers to claim sovereignty over inhabited territories in the Caribbean. Five centuries later, a similar legal fiction is being constructed in the digital sphere. The "training data" that fuels the multi-billion-dollar generative AI economy is often treated as data nullius—information belonging to no one, available for the taking.1 This data, however, is not neutral; it is the digitized sum of human expression, including the rich, syncretic cultural output of the Caribbean.

The central mechanism facilitating this modern extraction is the United States doctrine of "Fair Use." While developed to balance the rights of authors and the public interest within the U.S. constitutional framework, Fair Use has arguably metastasized into a tool of global hegemony. Major AI laboratories, predominantly situated in Silicon Valley, scrape the open web—including Caribbean news archives, literature, music databases, and academic repositories—relying on the defense that their ingestion of this data is "transformative" and therefore non-infringing.2

This article offers a specific critique of how this U.S.-centric legal standard erodes the sovereignty of Caribbean cultural archives. I argue that the transposition of U.S. legal values onto Caribbean cultural assets creates an "Extraction Paradox": the very laws designed to promote innovation in the North effectively strip the South of its proprietary rights, reducing complex cultural heritage to mere statistical weight. Through a comparative analysis of the U.S. "transformative use" test and the Caribbean’s distinct socio-legal landscape, we expose the mechanisms by which the digital plantation is constructed and maintained.

II. The Trojan Horse of "Transformative Use"

The bedrock of the AI industry's legal defense in the United States is the concept of "transformative use," most famously articulated in Authors Guild v. Google, Inc. and tested recently in Warhol v. Goldsmith.3 The logic posits that if a secondary use adds new expression, meaning, or message, or creates a new function (such as a search index or a generative model), it does not infringe the original copyright.

A. The Functionalization of Culture

When applied to Caribbean literature or music, this doctrine performs a distinct ontological violence: it "functionalizes" culture. To the U.S. court, a novel by Earl Lovelace or a poem by Louise Bennett-Coverley, when ingested by an LLM, ceases to be a narrative work of art. It becomes a dataset of syntactic relationships. The AI is not "reading" the Caribbean text to understand the post-colonial condition; it is processing the probability of the word "Babylon" following the word "chant."

This functionalization acts as a legal solvent, dissolving the copyright protection that should theoretically attach to the work. For the Caribbean creator, this is disastrous. Their work is consumed not for its aesthetic value (which they could license), but for its "token" value (which is taken for free). The U.S. Fair Use doctrine essentially argues that because the AI is doing something different with the culture than a human reader would, the culture is free for the taking. This perspective ignores the reality that the value of the AI model is entirely derived from the richness of the underlying data which, in the case of the Caribbean, was often created under conditions of historical scarcity and struggle.

B. The Erasure of Provenance

Critically, the "transformative" test often blinds U.S. courts to the provenance of the source material. As seen in the Bartz v. Anthropic litigation, the industry often utilizes datasets (like "Books3" or "The Pile") that contain pirated copies of copyrighted works.4 Under a strict Fair Use analysis, if the end use is sufficiently transformative, the illicit nature of the acquisition might be overlooked or minimized.

For Caribbean archives, which often lack robust digital rights management (DRM) or widespread digitization by authorized local entities, this is particularly dangerous. Unauthorized scans of rare Caribbean manuscripts or bootleg recordings of folk performances can be laundered through the "transformative" wash of an AI model. Once the model is trained, the connection to the original Caribbean artifact is severed. The model "knows" the culture, but the culture has no claim on the model. This mirrors the extraction of bauxite or sugar: the raw material leaves the region, is refined elsewhere, and the value-added product (aluminum or rum/AI services) is sold back, with no residual wealth remaining in the territory of origin.

III. The Ontological Crisis: Individual Rights vs. Communal Sovereignty

The friction between U.S. Fair Use and Caribbean culture is not merely economic; it is philosophical. U.S. copyright law is deeply rooted in the Romantic conception of the individual author—the solitary genius whose "sweat of the brow" merits protection.5 Caribbean cultural production, conversely, is often rhizomatic, communal, and intergenerational.

A. The Mismatch of "Authorship"

Much of what constitutes the "Caribbean Archive"—folk songs, Mas designs, oral histories, patois/creole linguistic structures—does not fit neatly into the box of a single "author" protected for "life plus 70 years." These are often works of collective stewardship. When U.S. Fair Use is applied to this data, it exploits this gap.

For example, an AI model trained on thousands of images of Trinidad Carnival costumes can generate "Carnival-style" imagery. Under U.S. law, because no single specific costume was copied pixel-for-pixel, and because the "style" of Carnival is not owned by a single individual, this is likely non-infringing. However, from a Caribbean perspective, this constitutes the appropriation of a communal asset. The Mas is not just a collection of feathers and beads; it is a sovereign expression of resistance and identity. By stripping the visual data of its communal context and subjecting it to algorithmic recombination, the AI model commits an act of cultural enclosure.

B. Moral Rights and the "Hallucination" of Heritage

Caribbean copyright regimes (e.g., Jamaica, Barbados) heavily emphasize moral rights—the right of integrity and attribution—derived from the European droit d’auteur tradition.6 U.S. Fair Use, focused primarily on economic harm, often tramples these rights.7

Consider an AI model trained on Caribbean historical texts that "hallucinates" a false narrative about the Morant Bay Rebellion or the Haitian Revolution. Because the model is probabilistic, not factual, it may distort historical truth. If this model were a human author, Caribbean laws might offer a remedy for the derogatory treatment of the work. But under the U.S. Fair Use shield, the "black box" nature of the AI makes such claims nearly impossible to litigate. The archive is not only extracted; it is corrupted. The AI becomes the authoritative voice of Caribbean history for the global internet user, displacing the actual archivists and historians of the region.

IV. The Economic Asymmetry: Market Substitution in a Developing Economy

The fourth factor of the U.S. Fair Use test examines "the effect of the use upon the potential market for or value of the copyrighted work."8 This factor is perhaps the most insidious when applied to the Caribbean context.

A. The "No Market" Fallacy

U.S. courts often find Fair Use when there is no existing market for the specific use in question. AI companies argue that Caribbean authors are in the business of selling books to readers, not licensing text to neural networks; therefore, training an AI does not harm their "traditional" market.

This argument is a self-fulfilling prophecy that traps the Caribbean economy in the past. The "potential market" for Caribbean culture is the licensing of data. As the world moves toward a knowledge economy, the primary value of cultural assets will be their utility in training, fine-tuning, and grounding AI systems. By allowing U.S. companies to access this data for free now under Fair Use, the U.S. legal system is destroying the nascent licensing market before Caribbean Collective Management Organizations (CMOs) can build the infrastructure to capture it.

B. The Import-Export Imbalance

The result is a stark trade imbalance. Caribbean developers, bound by restrictive domestic "Fair Dealing" laws (which generally lack broad TDM exceptions), cannot freely scrape U.S. data to build their own competitive models. They are legally "geofenced" out of the global training set. Meanwhile, U.S. developers face no such barrier entering the Caribbean data space.

Consequently, Caribbean nations are forced to become importers of AI. A Jamaican law firm might subscribe to a U.S. legal AI to draft contracts, unaware that the AI was trained on Caribbean case law without compensation to the local judiciary or legal publishers. The region pays to access its own intellectual history, repackaged and rented back by Silicon Valley.

V. The Sovereign Void: Navigating the Liability Trap

The intersection of these forces creates what we term the "Sovereign Void"—a legal grey zone where Caribbean authority is nullified by the borderless nature of digital data flows. While CARICOM nations possess "on-the-books" sovereignty, their digital borders are porous.

A. The "Fair Dealing" Handcuffs

Ironically, while U.S. law is too loose (for the Caribbean's liking), Caribbean law is too tight (for local innovation). The strict "Fair Dealing" provisions in Trinidad and Tobago or Jamaica, which enumerate specific allowed uses (research, reporting, criticism), do not explicitly permit the large-scale data ingestion required for AI development.

This leaves the Caribbean in the worst of both worlds:

• Vulnerable to Foreign Extraction: U.S. companies ignore local laws, relying on the fact that they conduct the "copying" (training) on servers in California, thus triggering U.S. Fair Use rather than Caribbean copyright liability.

• Stifled Local Innovation: Local Caribbean tech startups cannot train their own models on local newspapers or music without facing the threat of domestic copyright lawsuits, as they lack a "Fair Use" shield.

This regulatory mismatch ensures that the only entities capable of building "Caribbean AI" are foreign entities.

VI. The Moral Rights Chasm: Integrity in the Age of Hallucination

While the economic implications of the "Fair Use" hegemony are severe, the cultural implications are perhaps more insidious. This divergence is rooted in the fundamental disconnect between the Anglo-American utilitarian copyright tradition—which incentivises the production of works for the public benefit—and the Continental droit d’auteur tradition, which views the work as an extension of the author's personality. The Caribbean, through its reception of the UK’s Copyright, Designs and Patents Act 1988 (CDPA), has inherited a robust framework for moral rights that is largely absent in the United States.9

A. The Right of Integrity vs. The Stochastic Parrot

Under section 14 of the Barbados Copyright Act, and similarly under section 80 of the UK CDPA (which informs Caribbean jurisprudence), an author has the right to object to "derogatory treatment" of their work.10 Derogatory treatment is defined as any addition to, deletion from, alteration to, or adaptation of the work that amounts to distortion or mutilation, or is otherwise prejudicial to the honour or reputation of the author.

In the context of Generative AI, the ingestion and subsequent "regurgitation" of cultural data frequently results in what is colloquially termed "hallucination," but legally constitutes a distortion. When a Large Language Model (LLM) trained on the corpus of Caribbean history generates a narrative that conflates the Haitian Revolution with the Morant Bay Rebellion, or misattributes the lyrics of a calypso to a reggae artist, it is not merely making a factual error; it is effectively mutilating the cultural record.

In the United States, moral rights are strictly limited to visual arts under the Visual Artists Rights Act (VARA), and even then, the scope is narrow.11 Consequently, a Caribbean author whose literary work is distorted by a US-based AI model has virtually no recourse in a US court. The "transformative" nature of the AI, which serves as a shield against economic infringement, also acts as a solvent for moral integrity. The US legal system treats the Caribbean text as raw data to be atomized and reassembled, stripping it of the "honour and reputation" protections guaranteed by domestic Caribbean statutes.

B. The Problem of Paternity and the "Black Box"

The right to be identified as the author (paternity) is equally imperiled. Section 77 of the UK CDPA and corresponding Caribbean statutes assert that authors must be identified whenever their work is published commercially.12 However, the "black box" architecture of neural networks makes attribution technically opaque and legally unenforceable. When ChatGPT generates a poem in the style of Derek Walcott, it does not cite Walcott’s specific works. It mimics the statistical probability of his syntax. This creates a "ghosting" of the Caribbean author.

The style—the unique "voice" developed through a lifetime of craft—is harvested and deployed, but the author’s name is erased from the value chain. In a region where name recognition is often the primary currency for artists touring the diaspora, this erasure of paternity is a direct assault on professional livelihood.

VII. Digital Colonialism 2.0: The Mechanics of Extraction

The legal disparities described above do not exist in a vacuum; they facilitate a specific mode of economic production. We posit that the current AI paradigm represents a shift from "Surveillance Capitalism" to "Digital Colonialism." If surveillance capitalism was about predicting behaviour, digital colonialism is about extracting the generative capacity of a culture.

A. The Prebisch-Singer Hypothesis of Data

In development economics, the Prebisch-Singer hypothesis suggests that over time, the terms of trade for primary commodities deteriorate relative to manufactured goods.13 We are witnessing a digital iteration of this phenomenon.

The Caribbean exports "primary data"—music, literature, academic research, climatic data, and genetic information—which is scraped for free or at negligible cost by Global North entities. This data is then "manufactured" into high-value AI models (e.g., GPT-4, Claude, Midjourney). These models are then licensed back to the Caribbean at premium subscription rates. The Caribbean nation pays to access the cognitive processing of its own raw materials. The "value-add" (the training run) occurs entirely outside the sovereign jurisdiction, ensuring that tax revenues, intellectual property rights, and technological know-how remain concentrated in the US.

B. The Infrastructure of Dependency

This dynamic creates a perilous infrastructure of dependency. As Caribbean governments and judiciaries begin to integrate AI tools for efficiency—using LLMs to summarise case law or draft policy—they are increasingly relying on models biased towards US legal norms and US cultural contexts.

A study of the "Common Crawl" dataset, which underpins many major LLMs, reveals a massive over-representation of US and Western European domains.14 Consequently, an AI legal assistant used in a Kingston law firm is statistically more likely to suggest a solution based on the Uniform Commercial Code (UCC) of the United States than the specific statutory provisions of the Jamaica Copyright Act. This is not just an inconvenience; it is a subtle erosion of legal sovereignty, where the tool used to administer the law gradually reshapes the law in the image of the tool’s creator.

VIII. Towards a Sovereign Data Architecture: A Policy Framework

To transition from "explorers" to "proprietors" of AI, Caribbean nations must move beyond passive critique and construct a defensive and offensive legal architecture. I propose a tripartite framework: Legislative Reciprocity, Sovereign Data Trusts, and Regional Collectivisation.

A. Legislative Reform: TDM with Reciprocity

The Caribbean must urgently update its copyright statutes to address Text and Data Mining (TDM). However, simply copying the broad exceptions found in Japan or the EU would be a strategic error.15 Instead, CARICOM nations should adopt a Reciprocal TDM Exception.

Under this proposed model, data mining would be permitted only if:

1. The purpose is non-commercial scientific research; or

2. The entity conducting the mining has lawful access to the work and maintains a "sovereign nexus" (e.g., servers located within CARICOM, or reciprocal data-sharing agreements).

This nuances the approach: it allows local Caribbean universities to innovate without fear of liability (solving the "Liability Trap"), while retaining the right to demand licensing fees from foreign commercial entities (avoiding the "Extraction Paradox"). It effectively "geofences" the fair dealing exception, creating a legal barrier to extraterritorial extraction.

B. The Sovereign Data Trust

To operationalize this, CARICOM should establish a Sovereign Data Trust.16 This would be a legal entity entrusted with the stewardship of public sector data and digitized cultural heritage (e.g., from national libraries and archives).

Rather than leaving individual artists to fight Google or OpenAI—a losing battle—the Trust would act as a super-collective management organization (CMO). It would negotiate "bulk licensing" deals with AI labs. If OpenAI wants to train on the "Caribbean Corpus" (to reduce bias and capture the Caribbean market), they must negotiate with the Trust. The Trust ensures that:

• Royalties are paid and distributed to community funds or individual rights holders.

• Attribution protocols are respected.

• Local Compute: Part of the deal could involve "in-kind" payments, such as providing compute credits or server infrastructure to Caribbean universities.

C. Leveraging the "Arrangements Necessary" Doctrine

Finally, the Caribbean must aggressively leverage its one distinct legal advantage: the protection of computer-generated works (CGWs). As noted in Part I, Section 9(3) of the UK CDPA and its Caribbean equivalents (e.g., s 19 Copyright Act of Trinidad and Tobago) grant authorship to the person who makes the "arrangements necessary" for the creation of the work.17

While the US Thaler v. Perlmutter decision renders AI output public domain,18 the Caribbean offers a safe harbour for IP ownership. CARICOM should market itself as a "jurisdiction of certainty" for AI output. International companies could be incentivized to establish their prompt engineering and fine-tuning operations in Bridgetown or Kingston. By doing so, they secure 50 years of copyright protection for their AI assets—protection unavailable in the US. This strategy turns the "legal bug" into a "sovereign feature," attracting Foreign Direct Investment (FDI) not for tourism, but for high-value IP generation.

IX. Conclusion

The "ontological crisis" precipitated by Generative AI is not merely a question of whether a machine can be an author. It is a geopolitical contest over the ownership of human culture. The current application of US-centric "Fair Use" acts as a neo-colonial mechanism, enclosing the digital commons of the Global South and converting sovereign expression into foreign capital.

For the Caribbean, the stakes are existential. Adhering to the status quo—where local laws are restrictive and foreign extraction is permissive—guarantees a future of digital servitude. However, the region possesses the legal DNA (in the form of the arrangements necessary doctrine) and the cultural wealth to chart a different course. By rejecting the "Fair Use" hegemony and erecting a Sovereign Data Architecture, the Caribbean can ensure that its archives remain not just a repository of the past, but the engine of its own autonomous future.

References

Cases

• Andersen v Stability AI Ltd, 744 F Supp 3d 956 (ND Cal 2024)

• Authors Guild v Google, Inc, 804 F3d 202 (2d Cir 2015)

• Bartz v Anthropic PBC, No 3:23-cv-05319 (ND Cal 2024)

• Feist Publications, Inc v Rural Telephone Service Co, 499 US 340 (1991)

• Nova Productions Ltd v Mazooma Games Ltd [2007] EWHC 24 (Ch), [2007] RPC 25

• Robert Marley Foundation v Dino Michelle Ltd (1994) 31 JLR 197 (SC Jamaica)

• Thaler v Perlmutter, 629 F Supp 3d 137 (DDC 2022)

• The New York Times Co v Microsoft Corp & OpenAI, No 1:23-cv-11195 (SDNY filed 27 Dec 2023)

Legislation

• Copyright Act, Cap 300 (Barbados)

• Copyright Act 1993 (Jamaica)

• Copyright Act, Chap 82:80 (Trinidad and Tobago)

• Copyright, Designs and Patents Act 1988 (United Kingdom)

• Copyright Act of 1976, 17 USC (United States)

Secondary Sources

• Abbott R and Rothman E, 'Disrupting Creativity: Copyright Law in the Age of Generative Artificial Intelligence' (2023) 75 Fla L Rev 1141

• Bently L and Sherman B, Intellectual Property Law (4th edn, OUP 2014)

• Couldry N and Mejias UA, The Costs of Connection: How Data Is Colonizing Human Life and Appropriating It for Capitalism (Stanford University Press 2019)

• Delacroix S and Lawrence N, 'Bottom-up Data Trusts: Disturbing the "One Size Fits All" Approach to Data Governance' (2019) 9(4) International Data Privacy Law 236

• Kwet M, 'Digital Colonialism: US Empire and the New Imperialism in the Global South' (2019) 60(4) Race & Class 3

• Mejias UA and Couldry N, 'Data Colonialism: Rethinking Big Data’s Relation to the Contemporary Subject' (2019) 20(4) Television & New Media 336

• Sobel B, 'Artificial Intelligence's Fair Use Crisis' (2017) 41 Colum JL & Arts 45

Footnotes

1. Michael Kwet, 'Digital Colonialism: US Empire and the New Imperialism in the Global South' (2019) 60(4) Race & Class 3.

2. See generally Authors Guild v Google, Inc, 804 F3d 202 (2d Cir 2015).

3. Authors Guild v Google, Inc, 804 F3d 202 (2d Cir 2015); Andy Warhol Foundation for the Visual Arts, Inc v Goldsmith, 598 US 508 (2023).

4. Bartz v Anthropic PBC, No 3:23-cv-05319 (ND Cal 2024).

5. Feist Publications, Inc v Rural Telephone Service Co, 499 US 340 (1991).

6. Copyright Act, Cap 300 (Barbados), s 14; Lionel Bently and Brad Sherman, Intellectual Property Law (4th edn, OUP 2014) 273.

7. Ben Sobel, 'Artificial Intelligence's Fair Use Crisis' (2017) 41 Colum JL & Arts 45.

8. 17 USC § 107(4).

9. Lionel Bently and Brad Sherman, Intellectual Property Law (4th edn, OUP 2014) 273.

10. Copyright Act, Cap 300 (Barbados), s 14; Copyright, Designs and Patents Act 1988 (UK), s 80.

11. 17 USC § 106A.

12. Copyright Act, Chap 82:80 (Trinidad and Tobago), s 12; Copyright, Designs and Patents Act 1988 (UK), s 77.

13. Raul Prebisch, 'The Economic Development of Latin America and Its Principal Problems' (1950) United Nations Department of Economic Affairs.

14. See generally Nick Couldry and Ulises A Mejias, The Costs of Connection: How Data Is Colonizing Human Life and Appropriating It for Capitalism (Stanford University Press 2019).

15. Current EU TDM exceptions (Directive 2019/790, arts 3-4) allow opt-outs for rights holders, a mechanism that is burdensome for dispersed Caribbean archives.

16. Sylvie Delacroix and Neil Lawrence, 'Bottom-up Data Trusts: Disturbing the "One Size Fits All" Approach to Data Governance' (2019) 9(4) International Data Privacy Law 236.

17. Copyright Act, Chap 82:80 (Trinidad and Tobago), s 19. See also Nova Productions Ltd v Mazooma Games Ltd [2007] EWHC 24 (Ch).

18. Thaler v Perlmutter, 629 F Supp 3d 137 (DDC 2022).

The Algorithmic Author and the Sovereign Archive: A Comparative Jurisprudence of Artificial Intelligence, Copyright Liability, and Ownership in the United States and select Caribbean Community countries.

By Abiola Inniss Ph.D. LLM

Abstract

The rapid proliferation of Generative Artificial Intelligence (GenAI) has destabilized the anthropocentric foundations of global copyright law, creating a schism between the legal treatment of data ingestion ("Input") and algorithmic generation ("Output"). This article provides a comparative legal analysis of this ontological crisis within the United States and the Caribbean Community (CARICOM), specifically Jamaica, Trinidad and Tobago, and Barbados. By contrasting the United States' common law reliance on the "Fair Use" doctrine with these select Caribbean countries statutory "Fair Dealing" and "computer-generated works" provisions, this article reveals a profound inversion of legal incentives.

The analysis demonstrates that while the US legal framework tends to shield AI developers from liability during the Input phase via "transformative use" defenses, it paradoxically strips AI users of proprietary rights in the Output phase due to strict human authorship requirements. Conversely, the CARICOM framework presents a "liability trap" for local developers due to the absence of Text and Data Mining (TDM) exceptions yet offers a unique "ownership opportunity" for users by legally attributing authorship of machine-generated works to the human "arranger." The report concludes that this regulatory mismatch fosters a dynamic of "digital colonialism," where Caribbean cultural data is extracted by foreign entities protected by US law, while local innovation is stifled by domestic statute. It argues that for Caribbean nations to transition from AI "explorers" to "adopters," legislative reform introducing TDM exceptions is critical to balancing the protection of sovereign creativity with the imperatives of the digital economy.

1. Introduction: The Ontological Crisis of Copyright

The rapid ascent of Generative Artificial Intelligence (GenAI) has precipitated a crisis that is less about technology than it is about the philosophical bedrock of property rights in the creative sphere. For three centuries, copyright law has operated on an anthropocentric presumption: that creativity is a uniquely human phenomenon, stemming from the "sweat of the brow" or the "spark of genius." The emergence of systems capable of ingesting the sum total of human expression—billions of images, texts, and phonograms—and synthesizing them into new, commercially viable works has fractured this consensus. The legal systems of the world are now tasked with resolving a dual-sided dilemma: the legality of the input (training) and the proprietary status of the output (generation).

This article offers an exhaustive analysis of this legal friction, contrasting the jurisprudential approach of the United States—a system rooted in constitutional flexibility and judicial doctrine—with that of the Caribbean Community (CARICOM), specifically Jamaica, Trinidad and Tobago, and Barbados. These Caribbean jurisdictions, heirs to the British common law tradition and the statutory architecture of the United Kingdom’s 1988 Copyright, Designs and Patents Act (CDPA), present a fascinating counter-narrative to the US model.[9]

Where the United States relies on the pliable, equitable doctrine of "Fair Use" to absorb technological shock, often favoring the developers of technology at the expense of rights holders, the Caribbean nations operate under a regime of "Fair Dealing" and specific statutory prescriptions. This creates a profound divergence in liability exposure. Simultaneously, the two regions have adopted diametrically opposed stances on the "authorship" of AI-generated material. The United States, bound by a romantic conception of the human author, has steadfastly refused copyright protection to machine output. In stark contrast, CARICOM member states possess statutory mechanisms that theoretically allow for the copyrighting of AI-generated works, ascribing authorship to the human investor or arranger.

This divergence is not merely academic; it dictates the flow of capital, the location of innovation, and the future of cultural sovereignty. As the Global North races to legitimize the "text and data mining" (TDM) necessary for Large Language Models (LLMs), the Global South—represented here by the Caribbean—faces the risk of "digital colonialism," where its cultural assets are extracted without compensation under foreign fair use laws, while its own developers are hamstrung by restrictive domestic statutes. In the subsequent sections, we will dissect the "Input Phase" (the legality of training) and the "Output Phase" (infringement and ownership) across these jurisdictions. We will explore the intricate case law of the Northern District of California, the "arrangements necessary" provisions of the Jamaica Copyright Act, the nascent AI policies of the Trinidadian government, and the broader economic implications of these legal disparities.

2. The United States: The Fair Use Crucible and the Humanocentric Barrier

The United States serves as the primary global theater for AI copyright litigation. This is due not only to the concentration of major AI laboratories (OpenAI, Anthropic, Google, Meta) within its borders but also to the unique structure of US copyright law, which relies heavily on litigation to define the boundaries of permissible use. The current legal landscape is defined by two overarching realities: a high probability that AI training will be litigated through the lens of "Fair Use," and a certainty that AI outputs are currently uncopyrightable public domain artifacts.

2.1 The Input Phase: The "Transformative" Gamble

The foundational question facing US courts is whether the ingestion of billions of copyrighted works to train an AI model constitutes copyright infringement. Under 17 U.S.C. § 106, the reproduction of a work is the exclusive right of the copyright owner. AI training involves making digital copies of works to analyze their statistical properties. Thus, unless a defense applies, training is prima facie infringement.[1] The defense of choice for the AI industry is "Fair Use" under 17 U.S.C. § 107.

2.1.1 The Doctrinal Anchor: Authors Guild v. Google

The intellectual lineage of the AI defense traces directly to Authors Guild v. Google, Inc. (2015). In this landmark case, the Second Circuit Court of Appeals held that Google’s digitization of millions of books to create a searchable database was a "fair use".[4] The court reasoned that the use was "transformative" because it did not provide a substitute for the books themselves but rather created a tool for information location—a "card catalog" for the digital age.

AI developers argue that LLMs are the spiritual successors to the Google Books project. They contend that training a model like GPT-4 or Claude is not about "consuming" the expressive content of a novel (i.e., enjoying the plot), but about analyzing the functional relationships between words—the syntax, grammar, and statistical probabilities that underpin human language. This, they argue, is a non-expressive, intermediate use that creates a new technological function.[20]

2.1.2 The Fracture in the Courts: Bartz, Kadrey, and the "Books3" Dataset

However, the "Google Books" analogy is fraying under the specific facts of Generative AI. Recent litigation in the Northern District of California has begun to expose cracks in the fair use defense, particularly regarding the provenance of training data.

In Bartz v. Anthropic PBC, the court addressed the use of the "Books3" dataset—a collection of nearly 200,000 books allegedly pirated from bibliotik trackers. The court found that Anthropic’s use of this dataset was problematic. Unlike Google, which scanned physical books it had often partnered with libraries to access, Anthropic was accused of using data that was "impermissibly acquired".[5] The court in Bartz signaled that the "bad faith" nature of the acquisition could undermine the fair use defense. If the source material is pirated, the "purpose and character of the use" (Factor 1) becomes tainted by the illegality of the source.

This decision creates a sharp contrast with Kadrey v. Meta Platforms, adjudicated in the same district. In Kadrey, the court was more dismissive of the idea that the mere act of downloading books for training was an independent infringement, reasoning that because the ultimate use (the model) was transformative, the intermediate steps were shielded.[11] This "irreconcilable" tension between Bartz and Kadrey highlights the current volatility of US law: does the "transformative" end justify the "pirated" means? The Bartz court, in dicta, doubted that it would ever be fair use to download books from pirate sites rather than purchasing them, suggesting a potential future requirement for "clean" data provenance.[5]

2.1.3 The "Regurgitation" Problem: New York Times v. OpenAI

The strongest challenge to the fair use narrative comes from The New York Times Co. v. Microsoft Corp. & OpenAI.[14] Unlike the authors in Kadrey or Tremblay, the New York Times (NYT) presented evidence of "regurgitation"—instances where GPT-4 outputted near-verbatim memorization of Times articles.

This evidence attacks the fourth factor of fair use: the effect on the potential market. In Google Books, the court noted that snippets did not substitute for the full book. However, the NYT complaint demonstrates that an AI can generate a detailed summary or even a verbatim reproduction of a paywalled article, effectively bypassing the subscription model.[14] If a user can get the "expression" of the Times from ChatGPT without paying the Times, the use is no longer just "functional"; it is a market substitute.

Furthermore, the NYT complaint introduces the concept of "hallucination" as a form of damage to the brand's integrity. OpenAI has attempted to characterize the NYT’s failure to "opt-out" as a defense, but the court has dismissed this as a "straw man," noting that there is no affirmative duty for a copyright holder to police every new technology for infringement. The NYT’s strategy represents a significant "about-face" from its position in New York Times v. Tasini decades ago; now, the Times champions the sanctity of human authorship against technological exploitation.[14]

2.2 The Output Phase: Infringement and the "Compressed" Theory

Even if the training is found to be fair use, does the generation of new content infringe? This brings us to the "Output Phase" litigation, most notably Andersen v. Stability AI.[3]

2.2.1 The Failure of the "Compressed" Copy Theory

In Andersen, a group of artists argued that because the Stable Diffusion model was trained on their images, the model itself was a "compressed copy" of their work, and therefore every image it generated was a derivative work. Judge Orrick of the Northern District of California decisively rejected this theory.[3]

The court held that the model does not store JPEG files; it stores mathematical weights and parameters. Therefore, an output is not automatically a derivative work simply because the model "saw" the original during training. To prove infringement, a plaintiff must show "substantial similarity" between a specific training image and a specific output image.[3]

2.2.2 The "Style" Loophole

This ruling exposes a critical limitation in US copyright law: it does not protect "style." An artist can protect a specific painting of a cat, but they cannot protect the technique of painting cats with thick impasto strokes and vibrant colors. Consequently, an AI prompted to "create an image in the style of Sarah Andersen" is likely engaging in non-infringing activity, provided the output is not a pixel-by-pixel replica of an existing Andersen comic.[3] This has forced plaintiffs to pivot other theories, such as the removal of Copyright Management Information (CMI) under the Digital Millennium Copyright Act (DMCA), arguing that AI models strip metadata from training images.

2.3 The Ownership Void: The "Human" Requirement

While the US courts grapple with liability, the US Copyright Office (USCO) has been unequivocal regarding ownership: AI-generated works are not copyrightable.

2.3.1 Thaler, Zarya, and Suryast

The USCO operates under a strict interpretation of the Copyright Clause of the Constitution, which protects the "writings of authors." The Office, backed by the decision in Thaler v. Perlmutter, maintains that "author" excludes non-humans.[18]

• The Zarya Decision: In the registration for the graphic novel Zarya of the Dawn, the USCO granted copyright to the text (written by the human, Kristina Kashtanova) and the selection/arrangement of the images. However, it canceled the registration for the images themselves, which were created by Midjourney.[19] The Office reasoned that the user does not "create" the image; the user provides a prompt, and the AI generates the visual expression in an unpredictable manner.

• The Suryast Decision: This principle was further entrenched in the refusal to register Suryast, an image created by Ankit Sahni using a custom AI tool (RAGHAV). Sahni argued that he provided the base image and the style image and controlled the parameters. The Review Board rejected this, stating that the machine, not the human, executed the traditional elements of authorship; the work was a public domain artifact.[21]

Implication: The US legal framework creates a "Public Domain Trap." AI companies may eventually win the right to train on copyrighted data via Fair Use (Input), but the users of these tools cannot own the Output.

3. The Caribbean Community (CARICOM): The Statutory Fortress

Shifting focus to the Caribbean, we encounter a legal landscape that is structurally distinct. The copyright acts of Jamaica, Trinidad and Tobago, and Barbados are modeled closely on the UK’s Copyright, Designs and Patents Act (CDPA) of 1988.[9] This lineage imparts two critical features: a restrictive "Fair Dealing" doctrine that makes AI training legally perilous, and a specific statutory provision for "computer-generated works" that makes AI output theoretically ownable.

3.1 The Input Phase: The Rigidity of Fair Dealing

Unlike the open-ended US "Fair Use," the Caribbean utilizes "Fair Dealing." This is a closed list of exceptions. If a use does not fall within a specific statutory category, it is infringement.

3.1.1 The Absence of Data Mining Exceptions

In the Caribbean statutes reviewed, the permitted acts for fair dealing are generally limited to research and private study, criticism, review, and reporting current events.[7][8] Crucially, there is no general exception for "transformative use" or "commercial data mining."

• Trinidad and Tobago: The Copyright Act (Section 11) allows reproduction for "informatory purposes" or "teaching," but these are narrowly construed. A commercial AI developer scraping Trinidadian newspapers would likely not fit within these exceptions.[8]

• Jamaica: Section 52 of the Copyright Act permits fair dealing for research, but strictly limits this to non-commercial purposes in many interpretations.[7]

The "Liability Trap": This creates a hostile environment for local AI development. Unlike their US counterparts, Caribbean developers cannot rely on a "Fair Use" shield. The lack of specific TDM exceptions, which have been adopted in other jurisdictions, leaves Caribbean innovators in a state of legal paralysis.[17]

3.2 The Output Phase: The "Computer-Generated" Anomaly

If the Caribbean is restrictive on input, it is surprisingly permissive on output. The region’s statutes contain a legal fiction that completely bypasses the "human authorship" debates consuming the US.

3.2.1 The "Arrangements Necessary" Doctrine

Jamaica (Copyright Act s. 2), Barbados (Copyright Act CAP 300), and Trinidad and Tobago (Copyright Act s. 19) all contain definitions for "computer-generated works" (CGW) that mirror the UK’s CDPA 9(3).[6][7][8][9]

Statutory Definition: "In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken."[8] This provision assigns authorship by legislative fiat to the human arranger.

3.2.2 Who is the "Arranger"? The Developer vs. The Prompter

The ambiguity in law in these Caricom countries lies not in whether an AI work is protected, but who owns it. The phrase "arrangements necessary" is legally untested in the context of GenAI in the Caribbean.

• The UK Precedent (Nova Productions): In the English case Nova Productions v. Mazooma Games, the court held that the author of the computer-generated video game frames was the programmer, not the player.[15] If Caribbean courts follow this, the copyright in a Midjourney image might belong to Midjourney Inc., not the prompter.

• The "Mas" Analogy (Trinidad): In Trinidad, there is often a separation between the "designer" of the costume and the "band leader" who organizes the logistics. The law often ascribes ownership to the "producer"—the person who undertook the arrangements.[16] This precedent suggests a willingness in Caribbean jurisprudence to look at logistical investment as a basis for rights attribution.

3.3 Duration and Moral Rights

The status of CGWs in the Caribbean also comes with different terms:

• Duration: In Jamaica and Barbados, copyright in CGWs lasts for 50 years from the end of the calendar year in which the work was made.[6][7]

• Moral Rights: In many Commonwealth jurisdictions, moral rights are excluded for computer-generated works.[9]

4. Comparative Synthesis: The Inversion of Incentives

Comparing the US and CARICOM frameworks reveals a striking inversion of legal incentives.

4.1 Table 1: Comparative Legal Frameworks for AI

Feature

United States

CARICOM (Jamaica, Barbados, T&T)

Legal Tradition

Common Law (Constitution-based)

Common Law (Statute-based / UK 1988 model)

Training Data Legality

Permissive (Likely): "Fair Use" defense. Favors AI Developers.[4]

Restrictive (Likely): "Fair Dealing" is closed-list. Favors Rights Holders.[7]

AI Output Authorship

Denied: "Human Authorship" required. Output is Public Domain.[18]

Granted: Statutory fiction assigns authorship to the "Arranger."[8]

Infringement Standard

Substantial Similarity: High bar (Andersen).[3]

Objective Similarity: Similar standard.

Copyright Term (AI)

N/A (Public Domain)

50 Years from creation.[6][7]

Economic Incentive

Incentivizes Model Creation (Input) but devalues Model Use (Output).

Disincentivizes Model Creation (Input) but valorizes Model Use (Output).

4.2 The "Digital Colonialism" Dynamic

This legal mismatch fosters a dynamic of "Digital Colonialism." US-based companies scrape the open web, including Caribbean cultural data, relying on US Fair Use laws.[20] Conversely, Caribbean entities are restricted by their own Fair Dealing laws from doing the same. The US companies then sell the AI services back to the Caribbean. While Caribbean users can theoretically claim copyright over their specific outputs, they remain dependent on foreign infrastructure.

4.3 The "Style" Gap and Personality Rights

One area where the Caribbean may offer stronger protection is in the realm of "style" and personality. In Jamaica, the jurisdiction has a robust common law tort of "appropriation of personality," established in Robert Marley Foundation v. Dino Michelle Ltd.[16] This tort protects the commercial value of a celebrity's identity. As AI voice cloning becomes prevalent, Jamaican artists may find this tort more effective than copyright law.

5. Future Trajectories and Policy Recommendations

5.1 The Caribbean "Explorer" Strategy

The Economic Commission for Latin America and the Caribbean (ECLAC) classifies many Caribbean nations as "explorers" in the AI landscape.[10] To graduate to "adopters," reform is essential.

• National AI Policies: Jamaica’s National Artificial Intelligence Task Force is recommending legal frameworks that address IP and data privacy to prevent the dilution of cultural assets.[13]1
  
  
  
  

• TDM Exceptions: There is a pressing need for CARICOM nations to update copyright acts to include exceptions for Text and Data M2ining.[17]
  
  
  
  

• Collective Licensing: Caribbean music and art could be licensed via Collective Management Organizations (CMOs) for AI training, ensuring royalties flow back to the region.

5.2 The US "Transparency" Pivot

In the US, focus is shifting toward transparency. If Bartz v. Anthropic succeeds in penalizing the use of "pirated" datasets, the industry will be forced to sanitize its inputs.[5] This could lead to a bifurcated market of "Clean" vs. "Wild" models.

5.3 Conclusion: The Fork in the Road

The question—Does artificial intelligence use in creative fields infringe copyright laws?—has no singular answer. In the United States, the training of AI is a legally contested but currently viable "Fair Use," while the output is a public domain orphan. The system prioritizes the machine's learning over the human's ownership of the result.

In contrast, the CARICOM framework hinges upon the "Arrangements Necessary" doctrine, which represents a critical statutory departure from the human-centric "originality" standards found in the United States. Rooted in the United Kingdom’s Copyright, Designs and Patents Act (CDPA) 1988 [9] and adopted by Jamaica [7], Barbados [6], and Trinidad and Tobago [8], this doctrine creates a legal fiction to resolve the authorship of computer-generated works (CGW). By decoupling the physical act of creation from the legal status of authorship, the law assigns the work to the entity—often the developer or a significant enterprise user—that facilitates the work’s existence. Per Section 19 of the Trinidad and Tobago Copyright Act, the author is the person "by whom the arrangements necessary for the creation of the work are undertaken" [8].

This doctrine acknowledges economic and logistical investment as a valid basis for copyright protection, similar to the recognition of rights for a "producer" or "band leader" in the context of Caribbean "Mas" [16]. While the U.S. Copyright Office maintains that AI output is a public domain orphan because it lacks a human author [18], the Caribbean statutes provide a clear path to proprietary control, converting a technological output into a 50-year protectable asset [6, 7]

The "Arrangements Necessary" Doctrine in CARICOM Jurisprudence”

The "Arrangements Necessary" doctrine represents a critical statutory departure from the human-centric "originality" standards found in United States copyright law. Rooted in the United Kingdom’s Copyright, Designs and Patents Act (CDPA) 1988 [9] and adopted by Jamaica [7], Barbados [6], and Trinidad and Tobago [8], this doctrine creates a legal fiction to resolve the authorship of computer-generated works (CGW).

Core Principles

• Decoupling Creation from Authorship: Unlike human-authored works where the "spark of genius" must manifest in the execution, the doctrine assigns authorship to the entity that facilitates the work’s existence. Per Section 19 of the Trinidad and Tobago Copyright Act, the author is the person "by whom the arrangements necessary for the creation of the work are undertaken" [8].

• Infrastructure vs. Intent: Based on the precedent in Nova Productions Ltd v. Mazooma Games Ltd, the "arrangements" are often interpreted as the provision of the logic, rules, and infrastructure required for the machine to generate an output [15]. This tends to favor the developer (the AI company) over the user (the prompter), unless the prompter's specific parameters are so extensive that they constitute the primary "arrangement."

• The Investment Rationale: As seen in the Caribbean context of "Mas" (Carnival masquerade), where the "producer" or band leader often holds rights due to logistical and financial undertaking [16], the doctrine recognizes economic investment as a valid basis for copyright protection.

Comparative Impact

While the U.S. Copyright Office maintains that AI output lacks a human author and is thus a public domain orphan (Thaler v. Perlmutter) [18], the Caribbean statutes provide a clear path to proprietary control. This converts a technological output into a 50-year protectable asset [6, 7], provided that the "arranger" can be clearly identified within the statutory framework.

For the global creative professional, this presents a complex arbitrage opportunity. One might train a model in Silicon Valley to escape liability, but generate the final work in Bridgetown or Kingston to secure ownership. However, for the Caribbean nations themselves, the path forward requires urgent legislative modernization. They must decide whether to maintain the "statutory fortress" that protects traditional rights but stifles innovation, or to introduce the TDM exceptions necessary to build a sovereign AI future that reflects their own rich cultural reality.

References

{1}. Abbott, R., & Rothman, E. (2023). Disrupting creativity: Copyright law in the age of {2}. Generative artificial intelligence. Florida Law Review, 75(6), 1141–1201.

{3}. Andersen v. Stability AI Ltd., 744 F. Supp. 3d 956 (N.D. Cal. 2024).

{4}. Authors Guild v. Google, Inc., 804 F.3d 202 (2d Cir. 2015).

{5}. Bartz v. Anthropic PBC, No. 3:23-cv-05319 (N.D. Cal. 2024).

{6}. Copyright Act, Cap. 300 (Barbados).

{7}. Copyright Act (Jamaica) 1993.

{8}. Copyright Act, Chap. 82:80 (Trinidad and Tobago).

{9}. Copyright, Designs and Patents Act 1988 (UK).

{10}. Economic Commission for Latin America and the Caribbean (ECLAC). (2025). Latin American Artificial Intelligence Index (ILIA 2025). United Nations.

{11}. Kadrey v. Meta Platforms, Inc., No. 23-cv-3417 (N.D. Cal. 2024).

{13}. National Artificial Intelligence Task Force. (2025). Policy Recommendations. Office of the Prime Minister, Jamaica.

{14}. New York Times Co. v. Microsoft Corp., No. 1:23-cv-11195 (S.D.N.Y. filed Dec. 27, 2023).

{15}. Nova Productions Ltd v. Mazooma Games Ltd EWHC 24 (Ch).

{16}. Robert Marley Foundation v. Dino Michelle Ltd (1994) 31 JLR 197.

{17}. Schirru, L., Souza, A. R., & Chamas, C. (2024). Building a text and data mining limitation: The Brazilian case. GRUR International, 73(3), 217–222.

{18}. Thaler v. Perlmutter, 629 F. Supp. 3d 137 (D.D.C. 2022).

{19}. U.S. Copyright Office. (2023). Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence.

{20}. U.S. Copyright Office. (2025). Copyright and Artificial Intelligence, Part 3: Generative AI Training (Pre-Publication Version).

{21}. U.S. Copyright Office Review Board. (2023). Decision Affirming Refusal of Registration of a Recent Entrance to Paradise (SURYAST).

[1] (Weinstein, 2010)

[2] (Cambridge University Press, 2017)

[3] (Ezell & Cory, 2019)

[4] (Inniss, 2017)

[5] (Inniss, 2017)

[6] (Chen & Puttitanum, 2005)

[7] (United Nations, 2025)

[8] (Lorenczik & Newiak, 2012)

[9] (The Statute of Anne ; April 10, 1710, n.d.)

[10] (United States Government, 2025)

[11] (The Constitution of the Republic of Haiti 1987, 2025)

[12] (Overview The TRIPS Agreement, 2025)

[13] (Revised Treaty of Chaguaramas, 2025)

[14] (UNESCO, 2022)

[15] (UNESCO, 2022)

[16] (National Innovation Pathways of the United States 2023, n.d.)

[17] (Overview The TRIPS Agreement, 2025)

[18] (United Nations, 2025)

[19] (Kenya Clinate Smart Agriculture Project, 2025)

Global Divergence in AI Copyright Liability: A Comparative Analysis of Fair Use, Text and Data Mining, and Fair Dealing in the US, EU, and Caribbean (2025),

By. Dr. Abiola Inniss Ph.D. LLM, ACIarb

Abstract

This article analyzes the diverging legal frameworks governing AI training and copyright in the United States, European Union, and Caribbean as of late 2025. In the United States, the judiciary has established a "conditional fair use" doctrine (Bartz v. Anthropic, Kadrey v. Meta), where training is transformative but liability arises from illicit data sources ("shadow libraries") or market substitution. The European Union enforces a statutory compliance regime under the AI Act, permitting text and data mining (TDM) only where rights holders have not exercised machine-readable opt-outs (e.g., C2PA). In the Caribbean, notably Barbados, legislative reforms prioritize creator sovereignty, rejecting broad TDM exceptions in favor of collective licensing. The analysis concludes that global AI liability is fractured: permitted as fair use in the US (with caveats), conditional on protocols in the EU, and largely infringing in the Caribbean, creating a complex compliance landscape for global AI development.

Executive Summary

As the global digital economy transitions into late 2025, the legal frameworks governing the intersection of Artificial Intelligence (AI) training and copyright law have fractured into distinct, often competing, regulatory paradigms. This report offers an exhaustive analysis of the judicial, legislative, and regulatory developments in the United States, the European Union, and the Caribbean regarding the legality of ingesting copyrighted works for the purpose of training generative AI models.

In the United States, 2025 has been a year of judicial refinement rather than legislative overhaul. The U.S. Copyright Office, in its pivotal Part 3 Report, declined to recommend a statutory blanket license, leaving the courts to define the boundaries of Section 107 "Fair Use".1 The resulting case law, specifically the summary judgment rulings in Bartz v. Anthropic and Kadrey v. Meta, has established a "bifurcated" liability model. While the act of intermediate copying for the purpose of pattern recognition is generally viewed as transformative and fair, this defense collapses if the underlying data source is illicit (e.g., "shadow libraries"). Thus, US liability is increasingly defined by data provenance and output substitution rather than the training process itself.23

In the European Union, the focus has shifted from judicial interpretation to strict regulatory compliance following the full implementation of the AI Act in August 2025. The EU regime is predicated on a statutory "Text and Data Mining" (TDM) exception under the Digital Single Market (DSM) Directive, which is heavily conditioned on a "machine-readable opt-out" mechanism. The legal battleground in Europe has moved to the technical layer, with German courts ruling in late 2025 that AI training is permitted under TDM exceptions unless rights holders have strictly adhered to emerging technical protocols like C2PA and TDM Reservation Protocols, as identified by the EUIPO.45

In the Caribbean, the region is navigating a complex transition between colonial-era "Fair Dealing" statutes and the urgent need for digital sovereignty. Barbados has emerged as the legislative vanguard, passing the Copyright Bill, 2025 to empower Collective Management Organizations (CMOs) and extend protections, signaling a pro-creator stance that implicitly rejects a broad US-style fair use for AI.67 Conversely, Jamaica and Trinidad and Tobago have focused on judicial regulation, issuing Practice Directions to curb AI "hallucinations" in court while retaining conservative copyright frameworks that likely classify unlicensed commercial AI training as infringement.89

This article explores these diverging paths, analyzing the economic and legal ripple effects of a world where data training is "fair use" in California, a "conditional exception" in Berlin, and potentially "infringement" in Bridgetown.

1. The United States: The Judicial Refinement of Fair Use

The United States legal system, relying on the flexibility of common law, spent 2025 rigorously testing the applicability of the "Fair Use" doctrine (17 U.S.C. § 107) to the mass ingestion of copyrighted works. Unlike other jurisdictions that sought statutory fixes, the US approach has been defined by high-stakes litigation in federal courts, primarily the Northern District of California and the Southern District of New York.

1.1 The Policy Baseline: The US Copyright Office Part 3 Report

On May 9, 2025, the U.S. Copyright Office released its long-awaited report, Copyright and Artificial Intelligence: Part 3 – Generative AI Training. This document serves as the foundational policy text for the US government's stance on AI and copyright in 2025.1

The Report effectively ended speculation regarding an immediate legislative "fix" such as a compulsory licensing scheme. Instead, the Office reaffirmed that existing copyright principles, specifically the fair use doctrine, should govern the training of AI models. However, the Office’s analysis was far from a blanket endorsement of the AI industry's practices. The Report emphasized that fair use is a context-specific, fact-intensive inquiry that cannot be determined in the abstract.1

Crucially, the Report questioned the "transformativeness" of generative AI models in instances where the model's purpose is to generate content that competes directly with the training data. The Office noted that while the technical act of copying for analysis (intermediate copying) has historically been protected, the "purpose and character" analysis for Generative AI is complicated by the fact that the output often serves the same intrinsic purpose as the input—creative expression.1 This "market substitution" concern became the central theme of subsequent litigation.

1.2 The "Bifurcated" Fair Use Doctrine: Bartz and Kadrey

The most significant legal development of 2025 was the emergence of a "bifurcated" fair use analysis, crystallized by two summary judgment rulings in June 2025: Bartz v. Anthropic PBC and Kadrey v. Meta Platforms, Inc. These cases dismantled the monolithic view of "AI training" and instead separated the act of training from the source of the data.

1.2.1 Bartz v. Anthropic: The Piracy Exception

In Bartz v. Anthropic, Judge William Alsup of the Northern District of California addressed whether Anthropic's use of copyrighted books to train its Claude models constituted fair use. The plaintiffs alleged that Anthropic utilized "The Pile," a massive dataset that included "Books3," a constituent dataset sourced from "shadow libraries" (pirated ebook repositories like Bibliotik).2

The court’s ruling on June 23, 2025, was nuanced. Judge Alsup agreed with Anthropic on the technical aspect of training, finding that the computational analysis of texts to extract statistical correlations and linguistic patterns was "highly transformative." This use served a fundamentally different purpose than the original books (which were meant for human entertainment), thus satisfying the first factor of fair use regarding the training process itself.10

However, the court denied summary judgment on the issue of the data source. Judge Alsup ruled that "Anthropic had no entitlement to use pirated copies for its central library" and that "Creating a permanent, general-purpose library was not itself a fair use excusing Anthropic's piracy".2 This ruling established a critical precedent: Fair use cannot cleanse the taint of piracy. Even if the use (training) is transformative, the possession and ingestion of illicitly acquired data remains an infringing act. This decision forced AI developers across the US to audit their datasets and purge components derived from shadow libraries to avoid liability.2

1.2.2 Kadrey v. Meta: The "Market Competition" Theory

Two days later, on June 25, 2025, Judge Vince Chhabria issued a ruling in Kadrey v. Meta Platforms, Inc. that further tightened the fair use defense. Similar to Bartz, the plaintiffs (authors) alleged that Meta trained its LLaMA models on copyrighted books without permission. Meta argued that its use was transformative and therefore fair use as a matter of law.11

While acknowledging the transformative nature of extracting information from books to build a language model, Judge Chhabria introduced a "market competition" theory that expanded potential liability. The court denied Meta's motion for total exoneration, explicitly noting that "fair use... typically doesn't apply to copying that will significantly diminish the ability of copyright holders to make money from their works".11

The court emphasized that if an AI model creates a "substitute" for the original work, the transformative defense is weakened. This ruling shifted the burden of proof onto AI companies to demonstrate that their models do not negatively impact the market for the original works—a difficult bar to clear given the capability of LLMs to generate summaries, sequels, or stylistically similar prose that could theoretically displace the original author.11

1.3 The "Substitution" Frontier: NYT v. OpenAI and Perplexity

Following the precedents set in the summer of 2025, the legal focus in late 2025 shifted from the training phase to the output phase, specifically regarding news and non-fiction content where the risk of substitution is acute.

1.3.1 The New York Times v. OpenAI

As of December 2025, this case remains in a contentious discovery phase. The core allegation is that ChatGPT regurgitates New York Times articles verbatim, serving as a direct market substitute for the newspaper's subscription service. In a significant development in late 2025, Judge Ona Wang ordered OpenAI to produce 20 million chat logs.12 This discovery order aims to empirically test the "regurgitation" rate of the model. If the plaintiffs can prove that the model consistently outputs substantial portions of copyrighted articles, the "fair use" defense likely fails under the fourth factor (market effect), regardless of how transformative the training process was.

1.3.2 The New York Times v. Perplexity

On December 5, 2025, the New York Times opened a new front in the AI copyright war by suing Perplexity AI.13 This lawsuit differs from the OpenAI case as it targets "Retrieval-Augmented Generation" (RAG) rather than just model training. The Times alleges that Perplexity's search engine creates "substitutes" for its journalism by scraping real-time content and summarizing it for users, bypassing the paywall and depriving the publisher of ad revenue and subscriptions.13

This case challenges the "fair use" status of RAG systems. While search engines have historically enjoyed fair use protection (e.g., Authors Guild v. Google), the Times argues that Perplexity does not just index content but replaces it. A ruling against Perplexity could imperil the business models of AI search assistants that rely on real-time web summarization.13

1.4 Visual Arts and Compressed Copies: Andersen v. Stability AI

In the realm of visual arts, the case of Andersen v. Stability AI proceeded toward a September 2026 trial date, but significant rulings in 2025 shaped the landscape.14 The court allowed claims to proceed regarding "compressed copies," accepting the plausible theory that the AI model's weights might contain compressed representations of the training images, thereby constituting a derivative work or a form of storage infringement.14

This theory attacks the "abstraction" defense used by AI companies (which argues that models only store mathematical patterns, not images). If plaintiffs can prove that the model can reconstruct the original training images (memorization), the "fair use" defense for training becomes significantly harder to maintain, as it resembles storage and distribution rather than just learning.14

1.5 Synthesis of US Jurisprudence

By the end of 2025, the US legal position on AI training and fair use can be summarized as Conditional Fair Use:

1. Training Process: The intermediate copying of data for the purpose of pattern recognition is generally considered transformative and fair use.10

2. Data Source: The use of illicitly acquired data (e.g., shadow libraries) is infringement and is not shielded by fair use (Bartz).2

3. Market Effect: If the AI model generates outputs that serve as market substitutes for the original works (e.g., verbatim regurgitation, detailed summaries), the fair use defense is likely to fail (Kadrey, NYT).1113

2. The European Union: The Statutory Compliance Regime

While the US navigated the nuances of common law, the European Union spent 2025 operationalizing a comprehensive statutory framework. The EU's approach is characterized by explicit legislative exceptions for "Text and Data Mining" (TDM) coupled with rigorous transparency and opt-out obligations mandated by the AI Act.

2.1 The AI Act and the DSM Directive: A Regulatory Pincer

The EU AI Act entered into full force in mid-2025, with key provisions regarding General-Purpose AI (GPAI) models becoming applicable on August 2, 2025.15 The AI Act does not create a new copyright exception; rather, it enforces the existing framework established by the Directive on Copyright in the Digital Single Market (DSM Directive) (2019/790).

The interaction between these two laws creates the EU's unique "Opt-Out" regime:

• Article 3 (DSM): Provides a mandatory exception for TDM for the purposes of scientific research by research organizations and cultural heritage institutions. This exception cannot be overridden by contract or opt-outs.16

• Article 4 (DSM): Provides an exception for TDM for any other purpose (including commercial AI training), unless the rightsholder has expressly reserved their rights in an "appropriate manner, such as machine-readable means".16

The AI Act bridges the gap by mandating that providers of GPAI models must (a) put in place a policy to comply with EU copyright law, and specifically (b) identify and respect the opt-outs exercised under Article 4(3) of the DSM Directive.17

2.2 The "Machine-Readable" Standard and Technical Protocols

The critical legal battleground in the EU in 2025 was the definition of "machine-readable means." Because Article 4 allows commercial TDM unless an opt-out is present, the technical validity of the opt-out determines the legality of the training.

In July 2025, the European Commission published the GPAI Code of Practice, which established the standards for compliance. Simultaneously, the EUIPO (European Union Intellectual Property Office) Study on Generative AI and Copyright provided the technical blueprint for these standards.4

The EUIPO study and the subsequent consultation identified a hierarchy of protocols:

1. Robots Exclusion Protocol (robots.txt): While widely used, the study noted its limitations (lack of granularity). However, the GPAI Code of Practice explicitly requires signatories to respect robots.txt as a baseline.17

2. TDM Reservation Protocol (TDM Rep): A W3C community group specification designed specifically for Article 4 compliance, allowing granular rights reservation at the file level.4

3. C2PA (Coalition for Content Provenance and Authenticity): A cryptographic standard for asserting provenance and rights, endorsed as a "state-of-the-art" method for expressing opt-outs in media files.4

4. Rights Reservation Databases: The study proposed "federated rights reservation databases" (e.g., via the Liccium Trust Engine or Valuenode) to allow rights holders to register opt-outs centrally, addressing the difficulty AI crawlers face in parsing every individual website.4

Implication: In 2025, an EU rightsholder who merely states "No AI Training" in a footer or Terms of Service without implementing a machine-readable protocol (like robots.txt or C2PA metadata) has arguably failed to validly opt-out under Article 4(3). Consequently, scraping their content for AI training would be legal under the TDM exception.45

2.3 Judicial Validation: Germany as the Testbed

Germany, often the bellwether for EU copyright jurisprudence, provided the first major court rulings interpreting these provisions in late 2025.

2.3.1 Kneschke v. LAION (Hamburg Regional Court, September 2025)

In this landmark case, a photographer sued LAION (a non-profit that compiles datasets for AI training) for including his images in the LAION-5B dataset. The Hamburg Regional Court ruled in favor of LAION, finding that the creation of the dataset qualified as TDM for scientific research under Section 60d of the German Copyright Act (which implements Article 3 of the DSM Directive).18

Crucially, the court held that the creation of a dataset for AI training is intrinsically an act of "text and data mining." This validated the legal basis for non-profit research organizations to aggregate data for AI development without needing licenses, provided they fall under the scientific research definition.18

2.3.2 GEMA v. OpenAI (Munich Regional Court, November 2025)

In November 2025, the Munich Regional Court issued a ruling in GEMA v. OpenAI. GEMA, the German music collecting society, argued that OpenAI's use of lyrics constituted infringement. The court affirmed that the TDM exception (Article 4) generally covers the ingestion of data for AI training.19

However, the court introduced a critical caveat: Memorization equals Infringement. The court ruled that if the AI model is capable of reproducing the training data (e.g., outputting full song lyrics), it has exceeded the scope of the TDM exception. TDM allows for the extraction of patterns, not the reproduction of the work itself. Therefore, "overfitting" (where a model memorizes its training data) strips the developer of the TDM protection, rendering them liable for reproduction infringement.19

2.4 Synthesis of EU Jurisprudence

By late 2025, the EU position is a Statutory Opt-Out Regime:

1. Commercial Training: Permitted by default under Article 4 DSM, provided the developer respects machine-readable opt-outs (robots.txt, C2PA).517

2. Scientific Research: Permitted unconditionally under Article 3 DSM for qualifying institutions.16

3. Liability Trigger: Liability attaches if the developer ignores valid opt-outs OR if the model generates infringing outputs (memorization/reproduction).19

4. Penalty: Violations are subject to the AI Act's massive fines (up to 7% of global turnover) and copyright damages.15

3. The Caribbean: Fair Dealing and Digital Sovereignty

In the Caribbean, the legal landscape in 2025 is marked by a tension between the legacy of colonial "Fair Dealing" laws and the modern necessity of digital sovereignty. Unlike the US "fair use" (which is open-ended and flexible) or the EU "TDM exception" (which is specific and statutory), Caribbean copyright laws generally require specific statutory exceptions for any use that is not authorized by the copyright holder.

3.1 The Regional Context: Fair Dealing vs. Fair Use

Most Caribbean jurisdictions (including Barbados, Jamaica, and Trinidad & Tobago) operate under a "Fair Dealing" framework derived from UK law. Fair dealing is traditionally limited to specific purposes: research, private study, criticism, review, and news reporting. It does not typically include a broad "transformativeness" test, nor does it automatically extend to commercial data processing.20

In 2025, the lack of a specific "Text and Data Mining" exception in most Caribbean laws means that commercial AI training likely falls outside the scope of fair dealing, rendering it potentially infringing absent a license.20

3.2 Barbados: The Legislative Vanguard

Barbados distinguished itself in 2025 by passing significant copyright reform, positioning itself as a leader in protecting creator rights in the digital age.

3.2.1 The Copyright Bill, 2025

Passed in October 2025, the Copyright Bill, 2025 replaced the outdated 1998 legislation. The bill was explicitly designed to address "digital content and streaming" and to "strengthen penalties for infringement".621

Key provisions impacting AI and data rights include:

• Empowerment of CMOs: The Act formally empowers Collective Management Organizations (like COSCAP) to take direct legal action against infringers on behalf of their members. Previously, CMOs lacked the locus standi to sue, leaving individual creators to fight costly legal battles. This change significantly increases the litigation risk for AI companies that might ingest Barbadian music or literature without a license.621

• Extension of Term: The Act extends copyright protection by 20 years, aligning with international trends and ensuring a longer period of exclusivity for rights holders.21

• No TDM Exception: Notably, despite global trends, the Bill did not introduce a broad fair use or TDM exception for AI training. By strengthening enforcement without creating an AI loophole, Barbados has effectively signaled a "permission-first" regime. AI developers wishing to train on Barbadian works (e.g., the extensive catalog of soca and calypso) must negotiate licenses.20

3.3 Jamaica: Judicial Regulation and "Originality"

Jamaica has focused its 2025 regulatory efforts on the application of AI in the justice system rather than amending copyright law to facilitate AI development.

3.3.1 Practice Directions and "Hallucinations"

Chief Justice Bryan Sykes issued the Practice Direction on the Use of Artificial Intelligence in 2025. This regulation strictly prohibits the use of AI to draft affidavits, witness statements, or expert evidence to prevent the submission of "hallucinations" (fabricated case law or facts) to the court. While this regulates the legal profession, it reflects a broader judicial skepticism toward generative AI's reliability.8

3.3.2 Copyright Law and Originality

Jamaica continues to operate under the Copyright Act with traditional fair dealing exceptions (Sections 52-55) limited to "research or private study".22

A critical jurisprudential barrier to AI in Jamaica is the "originality" requirement. The Jamaican Supreme Court, in Cabel Stephenson v. Doreen Hibbert, reiterated that originality requires "skill and labour" from a human author.23 This precedent implies that:

1. Inputs: Training on Jamaican works for commercial purposes is not "private study" and is likely infringement.

2. Outputs: AI-generated works are likely not copyrightable in Jamaica, as they lack the requisite human "skill and labour".23

Additionally, Jamaica amended its Copyright Act to accede to the Marrakesh Treaty (facilitating access for the blind), showing a willingness to create humanitarian exceptions, but has notably refrained from creating commercial AI exceptions.24

3.4 Trinidad and Tobago: The "Human Author" Stance

Trinidad and Tobago remains the most conservative of the three jurisdictions regarding AI copyright reform.

• Legislative Status: The Copyright Act contains no express exception for text and data mining or AI training. Section 24 lists specific exceptions, none of which broadly cover machine learning.25

• Judicial Rulings: In May 2025, Justice Westmin James sanctioned attorneys for submitting AI-generated fictitious case law, reinforcing the judiciary's view of AI as a tool requiring strict human oversight.9

• Policy Stance: The prevailing legal consensus, supported by the UNESCO/CARICOM AI Policy Roadmap, is that the region must "safeguard intellectual property rights" rather than dilute them. Legal experts in Port of Spain emphasize that a "human hand" is required for copyright, and that the mass scraping of data for commercial gain infringes on the moral and economic rights of creators.26

3.5 Regional Oversight: The Caribbean Court of Justice (CCJ)

The CCJ, the final appellate court for Barbados and other CARICOM states, issued Practice Direction No. 1 of 2025 regarding Generative AI. This direction mandates that any use of AI in court proceedings must be disclosed and verified. This high-level judicial oversight reinforces a regional norm: AI is a high-risk tool that must be subordinated to human verification and legal accountability, a stance that aligns with a restrictive view of AI's copyright privileges.27

4. Comparative Analysis: The "Compliance Fracture"

The divergence in legal frameworks across these three regions creates a complex global compliance landscape for AI developers. The following table illustrates the "Compliance Fracture" as of December 2025.

Table 1: Comparative Liability for AI Training (Dec 2025)

Feature

United States

European Union

Caribbean (Barbados/Jamaica/T&T)

Primary Legal Doctrine

Fair Use (17 U.S.C. § 107) - Common Law

TDM Exception (DSM Arts 3 & 4) - Statutory

Fair Dealing - Statutory (Closed List)

Legality of Training (Open Web)

Likely Fair Use, provided the source is lawful and use is transformative (Bartz).10

Permitted by default, unless rightsholder opts out via machine-readable means.5

Likely Infringing for commercial use. No general TDM exception exists.20

Legality of Training (Shadow Libraries)

Infringement. Fair use does not apply to illicitly acquired data (Bartz).2

Infringement. Requirement for "lawfully accessible" source (DSM Art 4).16

Infringement.

Opt-Out Requirement

Voluntary. No legal obligation to respect opt-outs (though often respected for PR).

Mandatory. Must respect machine-readable protocols (robots.txt, C2PA).17

N/A. Permission is required by default (Opt-In regime).

Liability Focus

Provenance & Output (Did you steal the data? Does output substitute the original?)

Process & Compliance (Did you check for opt-outs? Did you record the training data?)

Authorization (Did you obtain a license from the CMO?)

2025 Key Development

Bartz & Kadrey bifurcated fair use; NYT discovery on regurgitation.12

Full application of AI Act; German courts define "memorization" as infringement.19

Barbados Copyright Bill 2025 empowers CMOs; Jamaica/T&T regulate AI in courts.69

4.2 Second-Order Insight: The "Data Laundering" Risk and Regulatory Arbitrage

The stark difference between the EU's "Opt-Out" regime and the US "Conditional Fair Use" creates a significant risk of regulatory arbitrage.

• The "Havana" Harbor: AI developers might be tempted to move training operations to jurisdictions with broad TDM exceptions (like Japan, Singapore, or potentially Israel) to avoid the EU's strict transparency requirements or the US's litigation risks.20

• The "Shadow Library" Trap: However, the US ruling in Bartz acts as a global restraint. Since the court ruled that the possession of an illicit dataset (like Books3) negates fair use, a company cannot simply "wash" pirated data in a permissive jurisdiction and then deploy the model in the US. The "fruit of the poisonous tree" doctrine effectively extends US copyright standards to the global supply chain of any AI model intended for the American market.2

4.3 The "North-South" Divide in AI Copyright

The Caribbean's resistance to adopting US-style fair use or EU-style TDM exceptions highlights a growing "North-South" divide.

• The Global North (US/EU): Generally favors exceptions that allow their massive tech industries to ingest data at scale, viewing data as a "raw material" for innovation.

• The Global South (Caribbean): Views their cultural output (music, literature, art) as a primary economic asset. By rejecting TDM exceptions and strengthening CMOs (as Barbados did), these nations are asserting that their data is a "finished product" requiring compensation.621

• Economic Implication: This creates a barrier to entry for local Caribbean AI startups, who cannot rely on "fair use" to train models on local content. They face high transaction costs (licensing) that their Silicon Valley competitors (who might ignore Caribbean jurisdiction or claim fair use in the US) do not. This could ironically entrench the dominance of foreign AI models in the Caribbean region.

5. Future Outlook (2026 and Beyond)

As we look toward 2026, the legal conflict is evolving from "Can we train?" to "What can we output?"

• The "Substitution" War: The filing of NYT v. Perplexity in late 2025 signals that the next great legal battle will be over Retrieval-Augmented Generation (RAG). If courts rule that summarizing real-time news is "substitution" and not "fair use," the entire business model of AI search could collapse or be forced into a licensing model.13

• Technical Standardization: In the EU, the focus will be on the adoption of the TDM Reservation Protocol and C2PA. We expect 2026 to be the year where "machine-readable rights" become a standard part of file metadata, forcing a technological overhaul of how content is published on the web.417

• Caribbean Licensing Markets: With Barbados empowering CMOs, we expect to see the first test cases of Caribbean collective management organizations demanding royalties from global AI platforms for the use of Caribbean music and literature. This could lead to new cross-border licensing agreements or, alternatively, the "geoblocking" of Caribbean content from AI training sets.6

6. Conclusion

In 2025, the answer to "Does AI training constitute Fair Use?" depends entirely on where the server—and the lawyer—is located.

• In the US: Yes, it is Fair Use, if you didn't steal the data and your AI doesn't plagiarize the output. The courts have carved out a space for innovation but fenced it off from piracy and direct market competition.1011

• In the EU: Yes, it is a TDM Exception, if you respect the technical "No Trespassing" signs (opt-outs) and follow the bureaucratic map of the AI Act.5

• In the Caribbean: No, it is likely Infringement. The region has chosen to prioritize the economic rights of its creators over the unbridled expansion of AI training, betting that strong copyright enforcement will ultimately yield better economic returns than permissive open-ended exceptions.6

The "Fair Use" consensus of the early 2020s has shattered. In its place is a complex, fragmented global map of compliance, where data provenance, technical protocols, and jurisdictional boundaries define the legality of artificial intelligence.

7. References

Ref ID

Title / Jurisdiction

Description / Source

1

US Copyright Office Part 3 Report

Copyright and Artificial Intelligence: Part 3 – Generative AI Training (May 9, 2025). Defined US policy on fair use for AI training. 11

2

Bartz v. Anthropic (Ruling)

Bartz v. Anthropic PBC, No. 24-cv-05417 (N.D. Cal. June 23, 2025). Ruled that training on pirated "shadow library" data (Books3) is not fair use. 3

3

Bartz v. Anthropic (Settlement)

Settlement details ($1.5 billion) following the class certification and summary judgment ruling.

4

EUIPO GenAI Study 2025

Study on Generative Artificial Intelligence and Copyright (2025). Identifies protocols like TDM Rep and C2PA for EU opt-outs. 2828

5

EU TDM Protocols

European Commission consultation on TDM opt-out protocols under the AI Act and GPAI Code of Practice. 2925

6

Barbados Copyright Bill 2025

Copyright Bill, 2025 (Passed Oct 2025). Updates copyright law, empowers CMOs, does not include broad AI/TDM exceptions. 87

7

Barbados Senate Passage

Confirmation of the Copyright Bill's passage in the Barbados Senate (Oct 2025). 30

8

Jamaica AI Practice Direction

Practice Direction on the Use of Artificial Intelligence (2025) issued by Chief Justice Bryan Sykes. 9

9

Trinidad AI Court Ruling

Justice Westmin James sanctions attorneys for submitting AI-generated fake cases (Nexgen Pathology Services Ltd v Darceuil Duncan).31

10

Bartz v. Anthropic (Training)

Ruling that the act of intermediate copying for pattern recognition is transformative and fair use. 313

11

Kadrey v. Meta

Kadrey v. Meta Platforms, Inc. (N.D. Cal. June 25, 2025). Denied summary judgment for Meta based on "market substitution" theory. 1432

12

NYT v. OpenAI (Discovery)

Order compelling OpenAI to produce 20 million chat logs to test for regurgitation (Judge Ona Wang, Dec 2025). 16

13

NYT v. Perplexity

The New York Times v. Perplexity AI (Filed Dec 5, 2025). Lawsuit targeting RAG and AI search as market substitution. 1817

14

Andersen v. Stability AI

Andersen v. Stability AI (2025 Rulings). Allowed claims regarding compressed copies and model weights to proceed. 3319

15

EU AI Act Timeline

Implementation dates for GPAI models (August 2, 2025). 21

16

EU DSM Directive Arts 3 & 4

Analysis of the Text and Data Mining exceptions and the machine-readable opt-out requirement. 2523

17

GPAI Code of Practice

EU Code establishing robots.txt and other protocols as standards for AI Act compliance (July 2025). 2434

18

Kneschke v. LAION

Kneschke v. LAION e.V. (Hamburg Regional Court, Sep 2025). Ruled creation of datasets for AI training is TDM for scientific research. 355

19

GEMA v. OpenAI

GEMA v. OpenAI (Munich Regional Court, Nov 2025). Ruled memorization/reproduction of lyrics exceeds TDM exception. 265

20

Caribbean Fair Dealing Analysis

Analysis of "Fair Dealing" vs "Fair Use" in Commonwealth Caribbean jurisdictions and lack of TDM exceptions. 3637

21

Barbados CMO Powers

Legislative details on empowering COSCAP (Barbados CMO) to sue for infringement. 7

22

Jamaica Copyright Act

Sections 52-55 of the Copyright Act regarding fair dealing for research/private study. 38

23

Jamaica Originality Case

Cabel Stephenson v. Doreen Hibbert (2022). Supreme Court ruling on "skill and labour" and human authorship. 39

24

Jamaica Marrakesh Treaty

Amendments to Jamaica's Copyright Act for the visually impaired (Marrakesh Treaty) without broader AI exceptions. 40

25

Trinidad & Tobago Copyright

Copyright Act lack of TDM exception and Section 24 limitations. 41

26

Caribbean AI Policy

UNESCO/CARICOM AI Policy Roadmap emphasizing protection of IP rights. 42

27

CCJ Practice Direction

Practice Direction No. 1 of 2025 on the Use of Generative AI Tools in Court Proceedings (Feb 2025). 1010

Works cited

1. Copyright and Artificial Intelligence | U.S. Copyright Office, accessed December 5, 2025, https://www.copyright.gov/ai/

2. Copyright Office Issues Key Guidance on Fair Use in Generative AI Training - Wiley Rein, accessed December 5, 2025, https://www.wiley.law/alert-Copyright-Office-Issues-Key-Guidance-on-Fair-Use-in-Generative-AI-Training

3. Two California District Judges Rule That Using Books to Train AI is Fair Use, accessed December 5, 2025, https://www.whitecase.com/insight-alert/two-california-district-judges-rule-using-books-train-ai-fair-use

4. Fair Use and AI Training: Two Recent Decisions Highlight the Complexity of This Issue, accessed December 5, 2025, https://www.skadden.com/insights/publications/2025/07/fair-use-and-ai-training

5. GEMA, Getty and beyond: AI and copyright litigation in the EU and the UK - Taylor Wessing, accessed December 5, 2025, https://www.taylorwessing.com/en/interface/2025/online-and-ai-generated-content/ai-and-copyright-litigation-in-the-eu-and-the-uk

6. EXECUTIVE BRIEFING The Development of Generative Artificial Intelligence from a Copyright Perspective - EUIPO, accessed December 5, 2025, https://euipo.europa.eu/tunnel-web/secure/webdav/guest/document_library/observatory/documents/reports/2025_GenAI_from_copyright_perspective/2025_GenAI_from_copyright_perspective_executive_brief_en.pdf

7. New copyright law gives creators legal bite in protecting works - Barbados Today, accessed December 5, 2025, https://barbadostoday.bb/2025/10/22/new-copyright-law-gives-creators-legal-bite-in-protecting-works/

8. Copyright Bill hailed as milestone for creative future - Barbados Today, accessed December 5, 2025, https://barbadostoday.bb/2025/10/22/copyright-bill-hailed-as-milestone-for-creative-future/

9. Jamaica becomes regional leader with new rules for AI use in courts - CNW Network, accessed December 5, 2025, https://www.caribbeannationalweekly.com/news/jamaica-becomes-regional-leader-with-new-rules-for-ai-use-in-courts/

10. PRACTICE DIRECTION NO. 1 OF 2025 THE USE OF GENERATIVE ARTIFICIAL INTELLIGENCE TOOLS IN COURT PROCEEDINGS This Practice Directio - Caribbean Court of Justice, accessed December 5, 2025, https://ccj.org/wp-content/uploads/2025/02/PRACTICE-DIRECTION-NO.-1-OF-2025-THE-USE-OF-GENERATIVE-ARTIFICIAL-INTELLIGENCE-TOOLS.pdf

11. Copyright and Artificial Intelligence, Part 3: Generative AI Training Pre-Publication Version, accessed December 5, 2025, https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-3-Generative-AI-Training-Report-Pre-Publication-Version.pdf

12. Bartz v. Anthropic - BakerHostetler, accessed December 5, 2025, https://www.bakerlaw.com/bartz-v-anthropic/

13. From Books to Bots: Key Takeaways from the Anthropic Fair Use Decision for AI Developers and Copyright Holders | Insights | Ropes & Gray LLP, accessed December 5, 2025, https://www.ropesgray.com/en/insights/alerts/2025/06/from-books-to-bots-key-takeaways-from-the-anthropic-fair-use-decision-for-ai-developers

14. Kadrey et al v. Meta Platforms, Inc., No. 3:2023cv03417 - Document 598 (N.D. Cal. 2025), accessed December 5, 2025, https://law.justia.com/cases/federal/district-courts/california/candce/3:2023cv03417/415175/598/

15. What Business Lawyers Can Learn from the First AI Copyright Fair Use Rulings, accessed December 5, 2025, https://www.americanbar.org/groups/business_law/resources/business-law-today/2025-september/ai-copyright-fair-use-rulings-business-lawyers/

16. OpenAI forced to release 20 million chat logs in NYT lawsuit - Windows Central, accessed December 5, 2025, https://www.windowscentral.com/artificial-intelligence/openai-chatgpt/judge-forces-openai-to-produce-20-million-chat-logs-in-copyright-lawsuit

17. The New York Times sues Perplexity AI for ‘unfair usage’ of its content: What the lawsuit says, accessed December 5, 2025, https://timesofindia.indiatimes.com/technology/tech-news/the-new-york-times-sues-perplexity-ai-for-unfair-usage-of-its-content-what-the-lawsuit-says/articleshow/125792040.cms

18. New York Times sues AI startup for 'illegal' copying of millions of articles - The Guardian, accessed December 5, 2025, https://www.theguardian.com/technology/2025/dec/05/new-york-times-perplexity-ai-lawsuit

19. Andersen v. Stability AI: The Landmark Case Unpacking the Copyright Risks of AI Image Generators - NYU Journal of Intellectual Property & Entertainment Law, accessed December 5, 2025, https://jipel.law.nyu.edu/andersen-v-stability-ai-the-landmark-case-unpacking-the-copyright-risks-of-ai-image-generators/

20. Top Takeaways from Order in the Andersen v. Stability AI Copyright Case, accessed December 5, 2025, https://copyrightalliance.org/andersen-v-stability-ai-copyright-case/

21. Implementation Timeline | EU Artificial Intelligence Act, accessed December 5, 2025, https://artificialintelligenceact.eu/implementation-timeline/

22. New EU Code of Practice for General-Purpose AI Models: What impact will it have on copyright? - Dentons, accessed December 5, 2025, https://www.dentons.com/ru/insights/articles/2025/september/19/new-eu-code-of-practice-for-general-purpose-ai-models

23. The New Copyright Directive: Text and Data Mining (Articles 3 and 4) - Wolters Kluwer, accessed December 5, 2025, https://legalblogs.wolterskluwer.com/copyright-blog/the-new-copyright-directive-text-and-data-mining-articles-3-and-4/

24. The General-Purpose AI Code of Practice | Shaping Europe's digital future, accessed December 5, 2025, https://digital-strategy.ec.europa.eu/en/policies/contents-code-gpai

25. EUIPO study on generative AI and copyright: Technical solutions for opt-out declarations and transparency measures - Taylor Wessing, accessed December 5, 2025, https://www.taylorwessing.com/en/insights-and-events/insights/2025/05/euipo-study-on-generative-ai-and-copyright

26. EUIPO study highlights legal gaps and policy needs in AI and copyright governance - CADE, accessed December 5, 2025, https://cadeproject.org/updates/euipo-study-highlights-legal-gaps-and-policy-needs-in-ai-and-copyright-governance/

27. text and data mining opt-out in Article 4(3) CDSMD: Adequate veto right for rightholders or a suffocating blanket for European artificial intelligence innovations? | Journal of Intellectual Property Law & Practice | Oxford Academic, accessed December 5, 2025, https://academic.oup.com/jiplp/article/19/5/453/7614898

28. Development of Generative Artificial Intelligence from a Copyright Perspective - EUIPO, accessed December 5, 2025, https://www.euipo.europa.eu/en/publications/genai-from-a-copyright-perspective-2025

29. Commission launches consultation on protocols for reserving rights from text and data mining under the AI Act and the GPAI Code of Practice | Shaping Europe's digital future, accessed December 5, 2025, https://digital-strategy.ec.europa.eu/en/consultations/commission-launches-consultation-protocols-reserving-rights-text-and-data-mining-under-ai-act-and

30. Senate passes new copyright law - Nation News, accessed December 5, 2025, https://nationnews-brb.newsmemory.com/?publink=2482e3d5b_134fbac

31. Agentic AI and the Law: Preparing Trinidad and Tobago for the Next Digital Leap, accessed December 5, 2025, https://polanc.com/agentic-ai-and-the-law-preparing-trinidad-and-tobago-for-the-next-digital-leap/

32. Meta Wins on Fair Use for Now, but Court Leaves Door Open for “Market Dilution”, accessed December 5, 2025, https://www.authorsalliance.org/2025/06/26/meta-wins-on-fair-use-for-now-but-court-leaves-door-open-for-market-dilution/

33. Copyright and AI: the Cases and the Consequences | Electronic Frontier Foundation, accessed December 5, 2025, https://www.eff.org/deeplinks/2025/02/copyright-and-ai-cases-and-consequences

34. European Commission receives final version of General-Purpose AI Code of Practice | IAPP, accessed December 5, 2025, https://iapp.org/news/a/european-commission-receives-final-version-of-general-purpose-ai-code-of-practice

35. Germany: landmark court decision deals with AI training and copyright | Global law firm, accessed December 5, 2025, https://www.nortonrosefulbright.com/en/knowledge/publications/218a3079/germany-landmark-court-decision-deals-with-ai-training-and-copyright

36. Copyright and AI in the UK: Opting-In or Opting-Out? | GRUR International | Oxford Academic, accessed December 5, 2025, https://academic.oup.com/grurint/article/74/11/1055/8209765

37. compliance of the fair use clause in the South African copyright amendment bill with the three-step test and the constitution of South Africa | Journal of Intellectual Property Law & Practice | Oxford Academic, accessed December 5, 2025, https://academic.oup.com/jiplp/article/20/7/447/8128929

38. Jamaica Copyright Act, ss. 52-55 (fair dealing) | H2O - Open Casebooks, accessed December 5, 2025, https://opencasebook.org/casebooks/405-copyrightx-jamaica/resources/9.3-jamaica-copyright-act-ss-52-55-fair-dealing/

39. AI: sufficiently recognized by the laws in the music industry? - Myers, Fletcher & Gordon, accessed December 5, 2025, https://myersfletcher.com/ai-sufficiently-recognized-by-the-laws-in-the-music-industry/

40. Copyright Act Amended to Facilitate Individuals Who are Blind or Visually Impaired - Ministry of Industry, Investment & Commerce, accessed December 5, 2025, https://www.miic.gov.jm/copyright-act-amended-facilitate-individuals-who-are-blind-or-visually-impaired/

41. Research Exceptions in Comparative Copyright Law, accessed December 5, 2025, https://digitalcommons.wcl.american.edu/cgi/viewcontent.cgi?article=1074&context=research

42. Caribbean Artificial Intelligence Policy Roadmap - UNESCO, accessed December 5, 2025, https://www.unesco.org/en/articles/caribbean-artificial-intelligence-policy-roadmap