An investigation has revealed that companies marketing their products online are increasingly utilizing AI-generated influencers on social media platforms, raising concerns about the need for better transparency in advertising practices.
Reports indicate that brands are embracing AI-generated content designed to simulate authentic customer interactions, often without clearly indicating that the individuals portrayed are not real people. The Guardian's inquiry also discovered that some content creators producing AI influencer material are being required to sign non-disclosure agreements (NDAs), preventing them from discussing their work publicly.
Currently, there are no legal frameworks mandating companies to inform consumers when their advertising includes AI-generated individuals. However, in the European Union, provisions under the upcoming Artificial Intelligence Act set to take effect in August will require clear labeling of AI-generated or altered content—such as deepfake images, audio, and videos. Notably, these regulations will not be applicable in the UK.
Consumer advocacy group Which? has emphasized the importance of ensuring that shoppers are made aware when AI-generated influencers are featured in marketing campaigns, as opposed to actual individuals.
One instance is the Instagram presence of Once, a photo app that enables users to capture disposable camera-style images for events. According to an analysis by Reality Defenders, a cyber-security firm that specializes in detecting deepfakes, the brand has likely employed AI-generated influencers in its promotional content. In several Instagram videos, a bride expresses her satisfaction with the Once app during her wedding, declaring, “Everybody expected a no-phone wedding, so I gave them cameras instead.” The related post credits the app as “@oncefilmapp.”
Once did not respond to inquiries about this situation.
In another featured video, a seemingly AI-generated woman comments, “I could kiss the interior designer who showed me this,” while demonstrating the Maket app, which uses artificial intelligence for home design and planning. A spokesperson for Maket mentioned that AI influencers serve as one of various methods to pilot creative ideas and marketing strategies on a small scale, rather than forming a central aspect of their marketing approach.
Additionally, the fashion label Ashle, based in Dubai, posted an image on social media of a woman purportedly wearing their clothing at a restaurant but later deleted the photo due to the appearance of an extra finger on the model. In a statement to The Guardian, Ashle clarified that all pieces are genuine, handmade garments but acknowledged that AI was used for early marketing visuals during their launch phase. The images were removed because the designs are no longer part of their collection.
Lisa Barber, editor of Which? Tech, pointed out that a recent study on deepfakes revealed that a troubling 70% of participants struggled to accurately identify real from fake videos, indicating a significant risk of consumers being misled by AI-generated content and vulnerable to scams. She stressed the need for companies to be transparent regarding the use of AI in their advertising, especially when featuring AI-generated influencers.
The Advertising Standards Authority (ASA) stated that their existing regulations do not explicitly prohibit the posting of AI-generated promotional content without disclosure. An ASA spokesperson noted, “While there are no rules against this, the content must adhere to advertising standards and not be misleading or socially irresponsible.” The evaluator clarified that the evaluation of AI usage would focus on whether it leads to a misleading perception of the product.
Clarissa Mansbridge, a former celebrity manager and creator of AI influencers for brands, reported that businesses often seek her expertise in crafting hyperrealistic digital personas. She revealed that companies can commission her to create entirely new AI avatars, which can be utilized for user-generated content (UGC). For instance, a beauty brand may request a video featuring a lifelike, unnamed individual applying sunscreen by a pool, giving the impression of influencer content.
Mansbridge estimated that between 40% to 60% of content from major brands could be generated through AI, although many creators are bound by NDAs preventing them from discussing the use of AI. She noted that the shift to AI is partly influenced by brands seeking cost-effective solutions to high-quality photography without facing the complications surrounding human influencers.
“The UGC boom began about three years ago, when influencers cleverly recognized brands' demand for content resembling everyday people using their products,” she explained. “This style of authenticity is rooted in resonance rather than authorship, and AI offers brands a scalable means to realize this.”
Moreover, businesses actively market AI-generated content that mirrors genuine customer experiences. Leeds-based artist Zac Rossiter recounted being approached by a marketing agency that proposed creating an AI-generated unboxing video for his artworks. Although this form of content has gained popularity due to its perceived authenticity, Rossiter ultimately declined the offer, stating he would never partner with an agency that utilizes fabricated AI unboxing videos instead of real customer reactions.
