Trademarking Your Face: The New Frontier Against Deepfakes?

AI: An Acute Issue?

Trademarks traditionally serve a commercial function by allowing a person to distinguish their goods or services from those of others.^[1] However, amidst the rapid adoption of artificial intelligence (AI) technologies, and the resultant proliferation of deepfakes,^[2] could trademarks emerge as a potential safeguard against the unauthorized exploitation of one’s physical likeness? English television presenter Jeremy Clarkson recently filed an application to trademark his face, prompted by the circulation of an AI-generated video falsely depicting Clarkson and his former Top Gear co-stars promoting a fraudulent cryptocurrency scheme.^[3] Clarkson has insisted that his motivation is protective rather than egotistical.^[4] Whether such an application can succeed remains uncertain. Using trademark in this manner—for a purely protective, non-commercial purpose—would stretch the law of trademark beyond its core function.

Precedent

Clarkson’s attempt to trademark his face is not unprecedented. The European Union Intellectual Property Office (EUIPO) database indicates that over 80 photorealistic photos or computer-rendered faces, predominantly of models or athletes, currently receive trademark protection.^[5] In 2018, Dutch model Maartje Verhoef made history by successfully registering a figurative mark of her face with the EUIPO, securing protection for various commercial uses of her portrait.^[6] Since then, several other models have received trademark protection for their portraits, reflecting an increasing desire to trademark one’s face and a growing acceptance of facial likeness as registrable in Europe.^[7]

North America, however, has not followed this trajectory—at least not yet. In the early 2000s, Tiger Woods failed in a claim for alleged trademark infringement arising from the unauthorized use of his portrait.^[8] The Circuit Court held that images and likeness do not attract trademark protection because they do not perform the core trademark function of designation.^[9] The legal landscape in Canada appears similar, wherein the statutory framework likely poses insurmountable barriers to obtaining protection of one’s face, especially for the purpose of protection from deepfakes.

The Canadian Context

While the prospect of using trademark law as a shield against unauthorized deepfakes is intriguing idea, trademark law in Canada is ill-suited for this purpose. The Trademarks Act expressly prohibits the trademark of the portrait or signature of any individual who is living or who has died within the preceding thirty years.^[10] This provision seems to foreclose the possibility of trademarking one’s face, even for commercial purposes. 

Additional barriers prohibit registration for protective purposes. The Act requires that a mark be distinctive.^[11] This factor has proven to be a significant hurdle to registration in Europe. For example, Dutch model Puck Schrover’s application to trademark her portrait was initially rejected because, according to the EUIPO, “the face in question lacked such distinctive characteristics” to meet the threshold for distinctiveness.^[12] This requirement not only underscores the commercial aspect of trademarking one’s likeness but also implies that the registrability of one’s likeness may depend on the perceived uniqueness of one’s physical features! The “use it or lose it” requirement poses a similar challenge because a face would need to be used in a commercial capacity to gain and maintain protection under the Act.^[13] Consequently, using trademarking as a safeguard against AI is an option unavailable to the average person, who is neither publicly recognizable nor commercially reliant on their image. 

Unanswered Questions

Beyond the substantial hurdles associated with registering a portrait or likeness as a trademark, significant doctrinal questions arise concerning the scope and duration of such protection. Trademarks protect the use a mark within defined categories of commercial activity, leaving open the question of how, if at all, such protection would operate outside a commercial context. More specifically, would trademark protection cover any unauthorized use of a person’s likeness or only to unauthorized commercial exploitation?

Questions also arise regarding the temporal limits of such protection. While trademarks may, in principle, be renewed indefinitely, it is unclear whether the same approach would be appropriate for marks consisting of a person’s likeness. Should protection persist beyond an individual’s death, as it does in the case of copyright? Moreover, given the key requirement of distinctiveness, extending trademark to facial likeness might incentivize individuals to pursue or maintain unique appearances. Changes over time—whether through aging, cosmetic procedures, or incidents causing facial deformities—may further complicate the assessment of whether a protected mark remains recognizable or distinctive.

Although these questions remain unanswered, they would be critical considerations if in any effort to extend trademark law as a means of protecting individuals against the harms posed by deepfakes.

Possible Solutions

Some jurisdictions have already begun responding to this issue, attempting to curb the use of AI for nefarious purposes. The Danish government recently introduced amendments to its copyright legislation that would prohibit the non-consensual sharing of deepfakes and grant individuals copyright protection over their own likeness.^[14] These amendments aim to empower citizens to compel digital platforms to remove unauthorized deepfake content.^[15]

In British Columbia, provincial privacy legislation may have laid the groundwork for responding to deepfakes. For example, section 3 of British Columbia’s Privacy Act prohibits the unauthorized use of a person’s likeness for commercial purposes.^[16] The advantage of seeking protection under the Privacy Act is that, unlike trademark, there is no requirement the person seeking protection engage in commercial activity. Furthermore, British Columbia’s Intimate Images Protection Act includes a definition of “intimate images” have “been altered in any way”,^[17] thereby encompassing deepfake images within in its scope.

However, considered together, the legislation falls short in its failure to protect non-commercial, non-intimate deepfake images. To date, the Privacy Act does not appear to have been tested in court against deepfake images. So, whether these provisions offer sufficient protection—and whether they can keep pace with the escalating sophistication and volume of deepfake content—remains to be seen. 

Conclusion

As deepfake technology accelerates, the tension between existing legal regimes and emerging threats to personal identity could become increasingly apparent. Trademark laws were not designed to protect individuals from AI-driven impersonation, and their structural limitations make them poor tools for that task. While the European Union’s more flexible approach has allowed some individuals to register facial likenesses as trademarks, these successes hinge on commercial distinctiveness and public recognition—criteria unattainable for most people. Canada’s statutory barriers are even more restrictive, effectively closing the door on such strategies. Without modifying intellectual property rights to meet the challenges of digital identity in the AI era, meaningful protection against deepfakes will likely exist beyond trademark law. 

AI Disclosure: AI-assisted tools were used for proofreading and stylistic editing only.

---

^[1] Trademarks Act, RSC, 1985, c T-13 at s 1(1) [Trademarks Act].

^[2] Deepfakes refer to “media manipulations that are based on advanced AI, where images, voices, videos or text are digitally altered or fully generated by AI.” See Government of Canada, “Deepfakes: A Real Threat to a Canadian Future” (14 July 2025), online: <canada.ca/en/security-intelligence-service/corporate/publications/the-evolution-of-disinformation-a-deepfake-future/deepfakes-a-real-threat-to-a-canadian-future (online)>.

^[3] Asyia Iftikhar, “Jeremy Clarkson is trademarking his face” (27 November 2025), online: <msn.com/en-gb/cars/news/jeremy-clarkson-trademarks-his-own-face-in-stand-against-ai/ar-AA1Rgs3W>.

^[4] Ibid.

^[5] Barna Arnold Keserű, “Trademark protection for faces? A comprehensive analysis on the benefits and drawbacks of trademarks and the right to facial image” (2024) JIPITEC 15 (2024) 15:1 JIPITEC 88 at para 8 [Keserű].

^[6] Ibid at para 5. 

^[7] Ibid at para 7.

^[8] ETW Corporation v Jireh Publishing, Inc 332F3d 915 (6th Cir 2003) [ETW Corporation v Jireh Publishing].

^[9] Ibid.

^[10] Trademarks Act at s 9(1)(l).

^[11] Trademarks Act at ss 1(1), 18(1)(b), 37.

^[12] GIDE, “The Case of Puck Schrover’s Face as a Trademark” (24 February 2025), online: <gide.com/en/news-insights/the-case-of-puck-schrovers-face-as-a-trademark/>.

^[13] Trademarks Act at s 4.

^[14] James Brooks, “Denmark eyes new law to protect citizens from AI deepfakes” (6 November 2025), online: <ap.org/news-highlights/spotlights/2025/denmark-eyes-new-law-to-protect-citizens-from-ai-deepfakes>.

^[15] Ibid.

^[16] RBSC 1996, C 373 [Privacy Act].

^[17] SBC 2023, C 11 at s 1.