Celebrities have filed trademark applications for unique voice and gesture traits to be able to claim legal rights against AI generated replicas of themselves.
Celebrities including Taylor Swift and Matthew McConaughey are using a new way to legally fight AI generated copies of themselves. In order to do so, they have been filing trademark applications based on clips of their voice and/or distinctive phrases and motions. These filings are being analyzed by The Conversation (a non-profit news organization), Renner Otto (lawyer) and Gallium Law (another lawyer).
These filings mark a first step in expanding upon the normal scope of trademark laws. Trademark law is generally used to protect brands and other intellectual property that is commercially related. Copyright law protects original work; however, no law protects people's biometric characteristics, therefore trademark filings are attempting to protect the biometric characteristics and mannerisms that AI can replicate today with "startling accuracy," according to Gallium Law.
Moving beyond normal trademark boundaries.
Trademark law has traditionally protected individuals' identities through means such as right of publicity actions brought under state law. However, trademark law also allows public figures to establish "perimeter[s] of clear ownership" around their own identities by filing for trademarks that protect not only their catchphrases or logos but also their voice and body language, as explained in Renner Otto's article dated May 8, 2019.
Get the latest model rankings, product launches, and evaluation insights delivered to your inbox.
The use of trademark law in this context complements existing protections afforded under state right of publicity laws. One advantage of utilizing federal trademark law as opposed to state level remedies is that it provides greater recourse for protecting cross-jurisdictional harm caused by AI-generated content distributed throughout multiple states or jurisdictions.
Gaps in current law prompt new approaches.
The impetus behind the expansion of trademark law to include protections for public figure identity stems from a perceived lack of protection currently available. Both The Conversation and Tolerance.ca note that copyright law was "not built" to address the protection of identifiable identity features synthesized by AI.
Renner Otto notes that while the current generation of generative AI systems can synthesize voices and facial movements at will, there exists a critical need for aggressive IP strategies among public figures in response to what is characterized as an urgent issue. This issue arises because until recently, AI systems were unable to generate credible synthetic versions of real-world individuals. The filing of trademark applications is an effort to develop and assert jurisdiction over areas of public figure identity prior to wide-spread unauthorized AI usage.
Whether this new application of trademark law will ultimately prove successful in the courts remains unclear. Since the primary purpose of trademark law has historically been to provide consumers with reliable identification of products/brands, allowing trademark owners to prevent consumers from confusing one product/brand for another, the use of trademark law to prevent AI impersonation blurs lines between commercial branding and individual identity.