Dev Gangjee

Data is increasingly recognised as one of the most valuable resources of the 21st century. Its value derives from its use in machine learning algorithms, a form of artificial intelligence that finds useful patterns within data and learns from experience. This ever-increasing value has led to data being characterised as an asset. Property rights are a tried and tested legal response when it comes to regulating valuable assets. Do we therefore need a new intellectual property right for data? This chapter argues that we do not. It charts the rapid rise and recent demise of the EU data producer’s right (DPR). The DPR was proposed in 2017, to incentivise the creation, dissemination, and commercial utilisation of machine-generated data. Today it leaves a fading policy footprint in the EU, having succumbed to the compelling arguments ranged against it. However, a right of this nature continues to be debated in international policy discussions. After introducing data as a valuable resource (Section 2), the first contribution of this chapter is to analyse why the proposed DPR was unsuccessful, as a cautionary tale for others contemplating a similar model (Section 3). This obituary emphasises why data remains such a challenging res for intellectually property (IP) law. Its second contribution is to suggest that the EU is developing an alternative framework to private property, as a resource management model for data (Section 4). While the emerging approach prioritises data access rights, this emphasis does not go far enough. There are valuable lessons to be learned from constructed commons models, as an alternative regulatory approach to private property.

Trade mark registration systems exist to provide useful information. Registers tell us who owns what. Until recently, it was axiomatic that registers for marks were directed at human readers – an applicant for a trade mark, trade mark registry examiners, vigilant competitors, employees of search and watching agencies as well as the occasional judge. This list now has a new entrant. What are the implications for the registered trade mark ecosystem, when algorithms begin to efficiently and comprehensively read trade mark registers? This chapter outlines the adoption of AI-enabled similarity assessment technology by search agencies, trade mark registries and watching agencies. Building on recent improvements in semantic and image searches, these algorithms identify conflicts between marks at the registry level. They provide a heuristically helpful (upstream) snapshot of conflict risks, based on two dimensions of similarity: marks and goods. However this simplified assessment may unintentionally edge out the more complex multi-factor likelihood of confusion test in a wider range of situations, including trade mark infringement analysis. The limits of these algorithms must be borne in mind.

Geographical indications (GIs) signal provenance for regional products and function as valuable collective brands. They are increasingly protected by 'sui generis' or independent protection systems, which require the registration of a product specification. Over several decades the functions and features of these independent regimes have stabilized, forming recognizable patterns. This chapter unpacks what we mean by sui generis protection for GIs. It argues that sui generis GI registration and protection systems are important sites where GI theory meets GI practice. Traces of the distinctive, foundational normative commitments of GI protection can be seen in the architecture of sui generis GI protection.

This chapter makes the case for joined-up thinking when approaching non-traditional signs in trademark law. Over the past three decades, trade mark registration has moved from up-front exclusions for certain categories of signs (no shapes, no colours) towards cautious and incremental acceptance. However the policy concerns generated by the grant of legal monopolies in such signs remain equally relevant today. The grant of an abstract colour mark to one trader closes off a part of the colour spectrum to others. Can we therefore allow such signs in to the system while successfully managing the tensions generated by their admission? Responding to this challenge, this chapter explores two potential responses. First, when permitting such marks to be registered, should we correlate the mark as characterised at the time of registration—agreeably modest in its scope and ambitions—with the mark as deployed in an enforcement context, where it otherwise tends to be read more generously? Second, when it comes to regulating non-traditional marks, should we move beyond historic upstream solutions—in the form of exclusions from registrability—and proactively consider additional scope limitation mechanisms when applying infringement tests and defences? Drawing on a range of EU and US decisions across various categories of non-traditional marks, the chapter argues that both questions should be answered affirmatively. Section II reviews the manner in which non-traditional marks came to be accommodated within trademark registration systems. Section III focuses closely on the characterisation of the mark at the time of registration. Trademark registration calls for a non-traditional mark to be depicted or represented (always), described (often) and classified according to type (where possible). This characterisation has profound consequences, as the ongoing Louboutin (C-163/16) red-soled shoes litigation before the Court of Justice demonstrates. Once characterised, the mark is then channelled into the relevant stream of substantive examination analysis. Since characterisation matters, applicants have learned to adapt, in order to overcome obstacles to registration. However where characterisation techniques have been used to subvert substantive criteria, registries and courts have responded by overriding the applicant’s own preferred characterisation with an objective assessment of the mark’s content. Section IV outlines the importance of consciously connecting the scope of the mark as characterised for the purposes of registration with its scope for the purposes of infringement. Section V concludes.

For well over a century, the notion of a distinctive link between regional products and their places of origin has been articulated in the language of terroir. This type of causal relationship – where the physical geography factors within a region leave their distinctive traces upon the end product – is reflected in the definition of a geographical indication (GI) found in Article 22.1 of TRIPS. However this chapter focuses on the underdeveloped, yet increasingly utilised possibility of a link between product and place resting on a ‘reputation… that is essentially attributable to its geographic origin’. As GI subject matter expands beyond wines, spirits and agricultural products to encompass confectionaries, charcuterie, cheese and crafts, a deterministic ‘physical geography’ explanation loses traction. This chapter proposes that historical evidence can complement that of geography, allowing us to more meaningfully answer the question of when a product’s reputation can be said to be essentially attributable to its geographical origin.

This chapter provides a legal framework for deciding when a geographical indication (GI) has crossed over the line into generic usage, as the common name for a type of product (such as cheddar cheese). Genericide is a process and one that is highly controversial. While Art 24.6 of TRIPS has emerged as the international reference point, it is remarkably underdeveloped. The question of whether a designation has become the ‘common name’ for a type of product requires an empirically informed answer. However, when setting out the test for genericide, there are opportunities for normative inflections to be introduced. Section 2 sets out the history of this controversial area and the commercial stakes which influence legal determinations. Section 3 identifies the four main structural issues to be addressed (which regime to opt for; who bears the burden of proof; what is the threshold or tipping point for genericide; and how the factors should interact). Section 4 proceeds to analyse the categories of evidence, drawing on comparative experiences with such categories in operation. Here the perception of the target audience (consumers or the general public, as informed by trade and expert opinion) must be gauged against the contextual backdrop of market conditions, legislative or bureaucratic classifications of the term and the actions of the rights holder. Section 5 concludes with a review of the options for preventing generic use in sui generis GI regimes, while also suggesting the avenues for reviving terms that were formerly GIs but have been declared generic. The comparative analysis is offered as a practical resource for decision makers, while also serving as a reminder that the overarching enquiry should remain focused on the public perception of the term, as opposed to the competing commercial interests at stake.

This paper traces the emergence of a new res or object of protection within European trade mark law. Proprietary rights in trade marks have conventionally been premised upon the mark’s ability to communicate useful information; namely, indicating the commercial source of goods or services. Granting exclusive rights to control the use of a mark preserves its ability to reliably signal origin. Contemporary EU trade mark law goes further and protects the more expansive brand dimension associated with a successful trade mark. The Court of Justice of the European Union (ECJ) has enabled this by recognising not only the origin indication function of marks, but also their advertising, investment and communication functions as well. The brand is a remarkably elusive and protean, yet undeniably valuable, intangible. So what are the doctrinal tools and techniques available to courts, registrars and legal practitioners, enabling them to work with such elusive subject matter? What are the corresponding assumptions about brand creation and sustenance that reinforce these techniques? And can they be reconciled with recent conceptualisations of branding emerging from marketing and consumer studies research? Since brand protection is a controversial development within European trade mark law, this paper unpacks the manner in which the brand is conceived of within European legal doctrine. The ECJ situates branding within a one-way broadcast model, while contemporary marketing research emphasises that brand formation is dialogic and iterative. The ECJ’s approach to brand propertization is therefore not only inaccurate in presuming single author brand creation, but also deeply troubling since it marginalizes consumer agency and reinforces the exploitation of their immaterial labour through the instrumentality of trade mark law.

Historically, likelihood of confusion has been the core infringement test for trade mark law, and it remains the most frequently applied test in infringement actions by far. However, there are noticeable differences in how it is applied by the Court of Justice of the European Union (CJEU), the General Court, and national courts; and questionable outcomes when it is applied in novel situations. This book is the first comprehensive and systematic account of the confusion test within the harmonised European trade mark system. It considers how the test is applied by national trade mark registries across EU member states, by the European Union Intellectual Property Office, by national courts, and by the CJEU. It offers practical guidance, while also evaluating more recent developments such as initial-interest confusion, post-sale confusion, the infringement of non-traditional marks and consumer responses to uses of trade marks on the internet.