Greenlandic cultural communicator’s face used in fake knitwear ads without consent

A Greenlandic cultural communicator says her likeness was used in AI-generated advertisements for handmade knitwear without her permission, Danish broadcaster DR reports.

Josepha Kuitse Kunak Thomsen discovered in February that her face appeared on the website niviandaja.com, which sells sweaters marketed as hand-knitted “with love and patience” by two sisters in Greenland. The site features a woman named “Nivi” whose appearance closely resembles Thomsen’s, including distinctive facial tattoos, a gap-toothed smile, and a birthmark near her eye.

“I thought, my face has been stolen,” Thomsen told DR, adding that the resemblance was unmistakable.

The site’s owner, Jamal Ranj Hameed Rasul, claimed in an email that the images were “100 percent AI-generated using a legal tool and are not based on any real person.” After Thomsen’s complaint, the images were removed from the Danish-language version of the site but remain on the English version.

Thomsen initially dismissed the issue but later found that followers on her Instagram account—where she shares Greenlandic cultural content with over 26,000 users—had noticed the ads. One Canadian follower alerted her that their father had nearly purchased the sweaters, believing they were her work.

“At first, I thought it was just me affected, but then I realised how many people might have been scammed by this site,” she said. “Now it’s hurting my reputation as a cultural communicator. I don’t want to be known as the person selling fake sweaters.”

Thomsen has reported the case to police, contacted Denmark’s Consumer Ombudsman, and filed complaints with Meta and the e-commerce trustmark organisation e-mærket. She also demanded that Meta stop using her Instagram content to train AI models.

Legal experts note that Danish marketing law prohibits the use of a person’s likeness in advertisements without consent. Sten Schaumburg-Müller, a media law professor at the University of Southern Denmark, told TV Midtvest that the case could violate these rules, even if the images were AI-generated.

AI specialists confirm that generative models, trained on real human images, can inadvertently produce likenesses of living individuals. Stefan Sommer, a professor at the University of Copenhagen’s Department of Computer Science, explained that AI “takes a kind of average of all the training data it has seen,” making such resemblances possible.

Thomsen, who is based in Viborg, Denmark, works as an independent communicator of Greenlandic cultural heritage. She described the incident as part of a broader problem: “It’s shocking how easy it is to get away with scams like this.”