Does the FTC’s Recent Influencer Guidance Address Robots?

Those in the advertising and retail industries have likely heard about the Federal Trade Commission’s recent guidance entitled “Disclosures 101 for Social Media Influencers.” Since the document, directed towards social media influencers and the disclosure of endorsements, was released on November 5, 2019, players in this space have wondered about its implications for the FTC’s future enforcement tactics. However, it might be beneficial to also consider whether the FTC’s guidance documents relate to influencers who aren’t human.

The concept of Artificial Intelligence (AI) influencers came into existence before 2016, but it did not seem to gain momentum until 2018. Since then, these influencers’ popularity has rapidly risen. AI influencers exist entirely in a virtual world, and in some cases are controlled by the brand they represent, but in all other ways behave like real human influencers. Best known for promoting fashion and lifestyle brands, they have often found the largest following on Instagram. Recently, however, they have begun to gain fame on YouTube as well. Some of the most well-known AI influencers are Lil Miquela, Shudu, and SK-II’s Yumi.

What is especially interesting about AI influencers is the paradox they have seemingly created. At a time when people will pay to avoid ads, AI influencers are actually gaining followers, despite the fact that their audience knows the accounts are purely driven by marketing. Lil Miquela was even named one of Time magazine’s top 25 most influential people on the Internet in 2018. The new window AI influencers give businesses to reach people who might otherwise avoid their advertisements might induce more and more companies to engage in this mode of advertising. Not only do AI influencers have this wider audience-reaching advantage, but they are also much easier to control than a human influencer, making concerns like avoiding bad press and creating the “ideal” representation of a company all but disappear.

In light of the potential that AI influencers will be widely used in the near future, it might benefit companies to take another look at the FTC’s guidance documents and think about how they might relate to non-human influencers. The Disclosure 101 Guidance represents the first time the FTC has officially spoken to the public about endorsements since its revision of the Testimonial and Endorsement Guides in 2009, which it largely enforced against actors other than influencers. However, starting in 2017 the FTC increasingly shifted its attention towards influencers, beginning with warning letters and culminating in the recent guidance. The Disclosure 101 Guidance restates the policies of the 2009 revision. Three of its fundamental points are that influencers must disclose material connections with brands, that they must ensure their reviews honestly reflect their experience with products, and that they must avoid making claims advertisers themselves cannot substantiate.

These policies raise many questions as related to AI influencers. For example, whose connection with the advertiser is considered material and thus must be disclosed, that of the AI influencer or those controlling it behind the scenes? Does it matter if the AI influencer’s creators are paid for their “endorsement”? The FTC’s guidance documents indicate that the answer to the first question is those controlling the AI, and that the answer to the second is a resounding “yes.”

However, there are questions to which the FTC’s guidance documents do not seem to provide an answer. Whose “experience” with the product matters, that of the AI influencer or of the creators? When we say an AI influencer has an experience with a product, how can this even reflect an honest belief? Does it matter that we cannot see or evaluate the credibility of those behind the AI? What if consumers can’t even tell that an influencer is computer-generated? After all, Lil Miquela pretended to be human until she announced in 2018 that she was a robot. Finally, what happens when artificial intelligence advances and the AI influencers become more autonomous?

If the FTC’s guidance documents raise issues relevant to AI influencers (which seems to be the case), the same could be said for any non-human influencer. For example, if Doug the Pug promotes dog biscuits, does it matter if we could never truly know his opinion of the product? What would a material connection mean for a dog or cat?

Though the FTC has not specifically addressed AI (or non-human) influencers, an agency spokesperson told CNNMoney that "advertisers using CGI influencer posts should ensure that the posts are clearly identifiable as advertising." Unless and until the FTC issues further statements, only these words, the 2009 revision, and the Disclosure 101 Guidance exist to give companies direction as to how to use AI influencers without running afoul of the disclosure rules.

For now, and until some of the above questions are answered, companies should be particularly careful to be truthful when they use AI influencers to promote their products, both about the computer-generated nature of the influencer and about the sponsorship behind the advertisement. Moreover, they should use the same degree of oversight and careful contracting language that they would with a human influencer if they choose to engage with an AI influencer they do not control.