Fans Urge Taylor Swift to Improve Following Allegations of AI Use in Promotional Videos

“We are significantly losing the fight against common sense regarding the use of generative AI,” Schnitt states, noting that if the videos are indeed AI-generated and Swift offers an apology, it could serve as “a touchstone moment” in the resistance against the technology.
Lobo, who shared a post with the #SwiftiesAgainstAI hashtag, believes Swift will refrain from addressing the backlash. She suspects the pop star, regardless of her involvement with AI, will be cautious about using it in the future, fearing backlash from her fanbase. In contrast to the promotional content, Lobo’s post on X highlighted Swift’s 2017 lyric video for “Look What You Made Me Do,” crafted by a motion design studio. Many fans commented on Lobo’s post expressing nostalgia for the artistry and meticulousness present in Swift’s earlier lyric videos.
“Back in the day, when she wasn’t as monumental as she is now, she was meticulous enough to hire someone to produce something both beautiful and carefully crafted,” Lobo remarks. “I have a job that is endangered by AI, which completely overlooks the art and reduces it to a mere product.”
While it’s unknown which AI models were used in creating the promo videos, Colman from Reality Defender points out that some models are trained on non-copyrighted data, while others cross into more dubious areas. However, mainstream AI products from firms like OpenAI and Google are currently wrestling with making the training of their models on copyrighted content permissible under fair use, much to the dismay of artists losing paid opportunities to AI.
Colman mentions that current generative AI models, combined with a “strong prompt,” could produce the types of images featured in Swift’s promos in roughly two minutes. Many of these videos are generated using diffusion AI models, which create outputs similar to Sora, OpenAI’s video application that allows users to easily deepfake themselves.
Google hinted at Swift’s scavenger hunt via its official Instagram account, although it’s unclear if the promotional videos part of the challenge utilized Google’s AI features. Earlier this year, Google has been marketing a tool that transforms photos into brief AI-generated videos, with the latest version named Veo 3. If the Swift teasers aimed to encourage her fans to engage with Google’s AI toolkit, the strategy appears to have backfired. This audience might actually be one of the most expressive and least likely groups to adopt AI tools.
Most individuals involved in the backlash are “devoted fans,” Lobo notes, who simply “don’t want AI to encroach on what we perceive as a safe space.” As long as Swift remains silent, it will stay uncertain whether AI was involved at all.