Virtual influencers are a new type of social media influencers, deployed by commercial entities to endorse their goods and services across several social media platforms. Distinguished by their entirely fictitious nature, virtual influencers exhibit remarkably human-like attributes encompassing physical characteristics and personalities. The visual representation of existing virtual influencers is claimed to be facilitated by artificial intelligence, although it is uncertain if this technology is really used to generate their images. Virtual influencers promote products in a way that human influencers do, albeit lacking the essential authenticity required to provide a faithful recommendation. They are fake personalities; hence they cannot try the products they recommend, a characteristic inherent to their human counterparts. In addition, virtual influencer profiles are under complete control of commercial entities such as brands. The average consumer developed the legitimate expectation that influencers endorse products based on their personal experience and opinion, rather than just following instructions from brands. However, in case of virtual influencers, experience-based endorsements are replaced by the financial incentives of the entities behind them. The lack of a personal element makes virtual influencer advertisements inauthentic and incommensurable with consumer expectations. Problematically, the average consumer remains oblivious to the fictitious nature of virtual influencers, or worse, their orchestration by commercial brands. The glaring absence of transparency pertaining to these elements leads to a general deception of the consumer.

The protection of consumers against such practices is harmonised in the European Union through the Directive on Unfair Commercial Practices. This paper explores the application of the standards set by EU law on virtual influencers, which are fictional characters without a form of natural or legal personhood. It ascertains that the deployer of virtual influencers, in addition to the third-party seller of the endorsed product, is in some cases bound by the provisions of EU law in this respect. We establish that the deception of the consumer by highly anthropomorphised virtual influencer posts that do not disclose the ad, the virtual nature of the influencer, and the control of the trader, is per se an unfair market practice that is forbidden under EU law. In this paper, we also take a closer look at the liability distribution between the parties involved in the creation process of these influencers, alongside the respective sanctioning that may be imposed. By doing so, we uncover several hardships when it comes to the enforcement of EU law provisions against misleading content created by virtual influencers. In addition, the paper uncovers the obligations of social media platforms to assess the systemic risks arising from such profiles, while suggesting a couple of specific policy measures these platforms could take to strengthen the existing enforcement mechanisms of EU and national law.

This working paper is a draft of a chapter that will appear in J. KERCKAERT and S. GEIREGAT (eds.), Social Media Influencers and the #Law, Heverlee, LeA Uitgevers, 2024.