When Technology Becomes a Brand: Can Algorithms Acquire Trademark Significance

Introduction

Today’s consumers don’t encounter brands via logos and slogans/brands’ websites. Instead of encountering brands’ websites or seeing the logos/identifiers you would associate with a particular company/brand on their respective websites, you encounter brands via algorithm-generated experiences, which create the framing of choice, preference, and behaviour. In today’s digital marketplace, the most prominent and frequently cited point of engagement between the consumer and the corporation’s brand to the consumer’s brand is algorithm-based (search rankings, recommended feeds, automated pricing, content prioritized, etc.) and the repeated exposure to these algorithm-generated experiences will, over time, produce an expectation or confirmation of brand familiarity, brand reliability, and brand difference, without the benefits that come with using a visible identifier or logo.

While consumers may not always be able to name an algorithm as the entity to which they are referring to when making a statement of trust with respect to a particular platform, they develop a unique and identifiable level of trust in the platform based on the way the algorithm responds to their needs and/or interests.

The accumulation of this trust translates into a brand experience through an algorithm’s output of technological behaviour, often in spite of the lesser role of traditional trademarks. However, the legal framework surrounding trademarks has historically only included visible trademarks as a source identifier and has never provided protection for a source identifier that has been incorporated into a product or service via the means of an algorithm.

As a result, the legal framework governing the protection of trademark rights is ill-equipped to address a market in which brand recognition is primarily driven by measuring the performance of the brand rather than properly using traditional brands and trademarks. In practice, technology companies are already exploiting this change in the marketplace by promoting the features generated by algorithms as unique, proprietary and trustworthy.

Consumers perceive a platform with superior algorithmic competencies to be a good target for their repeated use. When competitors impose similar algorithmic results of behaviour without actually imitating the business name or brand name, consumers can become confused and yet there would not necessarily be a straightforward proving of trademark infringement. The courts and trademark authorities have been cautious about extending trademark protection over algorithmic behavior as this could create a monopoly on functional advantages. However, if no protection is provided, this could ignore the reality that in today's digital ecosystems, algorithms will become more frequently used as source identifiers.

This creates a significant gap between how consumers currently view brands as source identifiers and how trademark law currently protects brands. With technology becoming the main source of trust in experiences, there is a significant urgency for trademark law to reconsider the possibility that source identification may come from consistent technological behaviour itself.

Trademark Law at the Crossroads: Can Algorithmic Behaviour Function as a Source Identifier?

The foundation of trademark law has always been that an item can be identified as coming from certain producers by virtue of the marking used to identify them with those items. When trademark law was created in the past, brands were recognised by visible indicators: names, logos, packaging, and symbols. However, in today’s digital marketplace, it has become common for consumers to engage with businesses through the results of algorithms rather than through visible indicators of brands. As a result, search results, recommendations, and automated results have become the primary ways that consumers trust and gain knowledge about a business. The consumer also expects that any experience they have had on a platform will come from the same source based on the consistent behaviour of an algorithm over time. In India, the recognition of the ability to use non-traditional elements as trademarks; for instance, the use of sound marks and other non-conventional identifiers as trademarks has emerged in case law.

The Indian Supreme Court confirmed in Cadila Health Care Ltd v Cadila Pharmaceuticals Ltd that consumer perception of a mark is critical to evaluating whether the mark functions as a trademark. In applying these principles, if consumers are relying upon the consistent behaviours of algorithms to assist them in identifying and returning to a specific platform, then that behaviour can be argued to also serve as a source-identifying function.

Trademark law also prevents the monopolization of functionality by ensuring that businesses cannot monopolize elements necessary for competition. With algorithms, the challenge lies in the invisibility of the algorithm itself versus the functionality of the results, which impact the identity of a brand. Historically, courts have dismissed trademark claims based on functionality because they wish to maintain the distinction between what is protected by a trademark and what is protected by patents. There are concerns about the manner in which algorithms affect how consumers perceive brands; therefore, trademark protection for algorithms will depend on whether consumers perceive brands through their distinctiveness or through the superiority of the functionality of those algorithms. The uncertainty created between trademark law and algorithms creates a conflict for courts as they must balance the interests of competition policy with the changing nature of how brands are experienced in technology-driven markets.

Enforcement Realities and the Risk of Functional Monopoly in Algorithmic Branding

When algorithms start having an impact on a company's brand name, the next step is to go from looking at the theory to actually enforcing it in the marketplace. In contrast to conventional trademarks, algorithmic branding does not have a stable or visible way for companies to defend their trademark rights since there is no solid, verifiable way for them to prove the algorithm's existence or that their algorithm has been copied in court. Because there is no tangible point of reference for the algorithm, it is very difficult to determine if someone has infringed upon the trademark as it is traditionally done through visual, auditory, or structural similarity.

Often, instead of duplicating an algorithm line-by-line, competitors will copy the user experience, the response pattern, or the outcome, and the consumer will act in a similar way as they do with the original algorithm but the copycat will not go through the traditional methods to determine if they have infringed or not. This situation will result in the courts looking at whether the consumer is relying on the inaccurate source-identification that they are seeking, and they will also consider whether the consumer is relying on the similarity of the functionality that has resulted from the copy of the original algorithm they have just purchased. This portion of trademark law draws a bright line that prevents a trademark from providing an exclusive right to control the functionality of a product as opposed to the brand name of the product.

The functionality doctrine is at the heart of trademark protection against the establishment of back-door monopolies on technological efficiency. The output of an algorithm typically overlaps with the two edges that the law believes to delineate between trademark protections and anti-competitive activity because the algorithm output has value added by the consumer's trust and expectation in addition to the functional benefit. The Indian courts have consistently placed a high priority on protecting market competition under these types of situations, rejecting trademark claims that would result in the freezing of innovation through perpetual trademark rights.

In terms of enforcement, this judicial caution has increased the evidentiary burden of claimants wanting to show confusion instead of just similarity to the mark at issue. As a result, businesses will not be able to rely on trademark law alone to adequately address algorithmic imitation in rapidly evolving digital markets. To address this practical concern, enforcement strategies continue to include trademark claims, unfair competition claims, contracts, and internal controls in their approach. This layered approach recognizes that algorithmic brand value does have value in the marketplace; however, it must be protected within the structural boundaries of trademark law.

Conclusion

With the advent of rapidly changing technology-driven markets, algorithms are becoming not only an essential part of back-end processes, but are also now the primary means by which consumers recognise and trust brands, thereby impacting their interactions with those brands. This change is challenging the traditional foundation of trademark law; traditionally, trademark law focused on protecting consumers' recognition abilities (name and logo) and not their experiential consistency (how well two experiences conform and are therefore trustworthy). In the marketplace today, many consumers return to the same platform based on the outcome of their previous experience, rather than based on the name or logo of the brand involved. However, trademark law limits the ability of a trademark to provide the consumer with a functional advantage over competitors by restricting the scope of the protections granted to a trademark.

The traditional approach to balancing protecting source identification with protecting competition within innovation driven markets is based addressing the brand via contractual remedies and competitive positioning, rather than through trademark law. As such, the resulting gap has created a situation where there is recognition of brand value by the business; however there is insufficient legal recognition to support the brand value claimed by the business. Businesses are therefore now required to address their intellectual property strategy in a more sophisticated manner; specifically, understanding that no one form of intellectual property protection could address the brand's ability to take advantage of algorithmic branding in and of itself. Trademark law may ultimately serve as an adjunct to protect the consumer's associations with the brand; however it will not serve as a standalone solution to protect the consumer's association with the brand.

Now that courts, regulators and businesses are all facing how to apply established legal principles to new technological realities, the future of trademark law will be less about expanding traditional doctrine and more about evaluating consumer impressions and their effects in the market. While the future of commercial identity will be defined by algorithms, the true challenge is not attempting to fit algorithms into existing legal definitions, but rather recognizing how they change and grow beyond the current boundaries of trademark protection. As a result, the intersection of technology and branding requires a rethinking of trademark law with an emphasis on protecting and encouraging innovation while protecting fair competition. In this rethinking, the inquiry becomes not whether algorithms impact brand identity, but how can trademark law respond to that impact in a manner that does not compromise the core principles of trademark law.

Author: Priyanka Pathak, in case of any queries please contact/write back to us via email to chhavi@khuranaandkhurana.com or at  Khurana & Khurana, Advocates and IP Attorney.

References

1. Trade Marks Act 1999 (India)

2. Cadila Health Care Ltd v Cadila Pharmaceuticals Ltd (2001) 5 SCC 73 (Supreme Court of India)

3. World Intellectual Property Organization, Trademarks: What is a Trademark? https://www.wipo.int/trademarks/en

4. World Intellectual Property Organization, Non-Traditional Trademarks https://www.wipo.int/trademarks/en/mark_types

5. Dev Gangjee, Relocating the Law of Geographical Indications (Cambridge University Press 2012)

6. Lionel Bently, Brad Sherman, Intellectual Property Law (5th edn, Oxford University Press 2018)

7. Barton Beebe, ‘The Semiotic Analysis of Trademark Law’ (2004) 51 UCLA Law Review 621

8. Graeme B Dinwoodie and Mark D Janis, Trademarks and Unfair Competition: Law and Policy (5th edn, Wolters Kluwer 2018)

9. Stacey L Dogan and Mark A Lemley, ‘The Functionality Doctrine and Trademark Protection’ (2006) 17 Stanford Technology Law Review 1

10. European Union Intellectual Property Office, Guidelines for Examination of European Union Trade Marks – Absolute Grounds https://www.euipo.europa.eu