USA Trends and Developments Contributed by: Anne Shea Gaza, Adam W Poff, Pilar G Kraman and Robert M Vrana, Young Conaway Stargatt & Taylor, LLP
Trade mark law: infringement and dilution The US Lanham Act, which codifies federal trade mark law, aims to protect businesses and their customers by regulating the use of distinctive marks, ie, brand names, logos, and other identi - fiers that indicate the business behind the good or service and may distinguish such a product from others. As with copyrighted works, howev - er, a generative AI model may have been trained on data that includes numerous trade marks belonging to well-known businesses. AI models that are trained on existing datasets including trade marks then create new logos, names, or symbols that may create output that is confus - ingly similar to existing trade marks. Generative AI’s ability to generate large numbers of designs in a short amount of time increases the risk of accidental or intentional infringement. For exam - ple, if a generative AI model is used to create a logo for a business, its output could be too similar to the existing logo for another business because the existing logo was included in the dataset used to train the AI model. The new mark constitutes trade mark infringement if it is “con - fusingly similar” to the existing mark, such that a consumer of the goods or services in question would be likely to be confused as to their source. Likewise, the output of generative AI could potentially cause trade mark “dilution”. A famous trade mark can be diluted where a similar mark is used by another in a way that weakens its distinctiveness in the eyes of consumers. This is particularly the case when the reputation of the owner of the famous trade mark may be tarnished by association with inferior goods or services. The potential for generative AI models to produce false or undesirable results has been well-documented, and a generative AI model could easily produce an offensive or decep - tive output, while associating that output with a well-known, real trade mark. For example, an AI-
generated news story that was false, appearing with the name and byline of a real newspaper, could cause harm to the newspaper’s reputation for accurate reporting, thereby diluting the value of its name. The Supreme Court on copyright fair use The doctrine of “fair use” is a defence to both copyright infringement and trade mark infringe - ment, although it is slightly different in each con - text. It allows the use of a copyrighted work or a trade mark, without the owner’s permission, under certain circumstances. Recently, however, the US Supreme Court issued decisions narrow - ing such use under both copyright law and trade mark law. In May 2023, the Supreme Court decided The Andy Warhol Foundation v Goldsmith. The Court agreed with photographer Lynn Goldsmith that a design by Andy Warhol – “Orange Prince” based on a photograph of musician Prince – was not a fair use of the original photograph when the War - hol Foundation licensed the design for use as a magazine cover. The 7–2 majority found that the Warhol design shared the same commercial pur - pose as the original photograph by Goldsmith, so it was not a fair use of the copyrighted pho - tograph. Fair use of a copyrighted work requires an analysis of: • the purpose and character of use, sometimes referred to as the extent to which the use is “transformative”; • the nature of the copyrighted work; • the amount and substantiality of the portion taken; and • the effect of the use upon the potential mar - ket. The Warhol Court focused on the final prong by emphasising the commercial purpose of the use,
752 CHAMBERS.COM
Powered by FlippingBook