USA Trends and Developments Contributed by: Matthew Vittone, Frankfurt Kurnit Klein & Selz
ing that the character’s “experience” is real. Instead, advertisers should make clear disclosures of the brand relationship and ‒ where material to consumer percep - tion ‒ that the persona is AI-generated. In other words, when using virtual influencers, advertisers are obli - gated to ensure that consumers are not being misled into thinking that the synthetic character’s endorse - ment represents an authentic consumer experience. Fake reviews and testimonials The FTC published its Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, which is intended to target unfair and deceptive practices involving consumer reviews or testimonials. Among other things, the rule prohibits reviews and testimoni - als that falsely claim to be from a real person, includ - ing those generated by AI or persons who do not have actual experience with the company’s products or services. The rule also enables the FTC to seek civil penalties against knowing violators. Fake reviews and testimonials have also been target - ed as part of Operation AI Comply. In one example, the FTC brought a claim against an AI writing assistant advertised as a way to generate consumer reviews. The FTC barred the company from marketing services intended to fabricate testimonials, citing the risk of misleading consumers who rely on authentic reviews. The takeaway for advertisers is that if an AI tool can be used to create deceptive content at scale, regulators may view the tool itself as problematic. Chatbots Marketers offering customer service through an AI- powered chatbot should make clear to consumers that they are communicating with a machine where consumers would reasonably think they are chatting with a person or where non-disclosure would be mis - leading. The FTC has emphasised in blog posts that chatbots must not mislead consumers about what they are or what they can do. If consumers are likely to believe they are interacting with a human when in fact they are communicating with AI, the omission of that fact can be deceptive. Certain state laws also require disclosure when using bots to communicate with consumers. By way of example, the recently enacted California Bot Dis -
closure Law prohibits using bots with the intent to mislead consumers in connection with a consumer transaction. To comply, brands must provide clear and conspicuous disclosures and the placement of the disclosures must be such that a reasonable per - son interacting with the bot understands they are not speaking with a human. The law applies to anyone using a bot to communicate or interact with a person in California online. When using AI-powered chatbots, advertisers are also responsible for the truth and accuracy of any claims made by their chatbots. And, if marketers should use AI tools to trick or manipulate consumers into mak - ing harmful choices that are contrary to their intended goals (so-called dark patterns), that may be consid - ered an unfair practice. Bias Even apparently “neutral” technology can produce results that are biased or discriminatory. To avoid inad - vertently introducing bias or other unfair outcomes, the FTC advises users of AI to scrutinise algorithms and data sets for biases, and to proactively monitor AI outputs and embrace transparency. Digital replicas Advertisers are increasingly harnessing the use of digital replicas. These are computer-generated sim - ulations of a real person’s image, likeness, voice or performance that is virtually indistinguishable from the original. The rapid rise of AI-generated image, voice and like - ness “clones” has pushed US law-makers and regula - tors to confront the risks of digital replicas and deep - fakes. In general, while no federal right of publicity law exists in the USA, state right of publicity laws protect individuals’ identities (name, image, likeness, etc) from unauthorised use ‒ examples of which include the fol - lowing. • Tennessee’s ELVIS (Ensuring Likeness Voice and Image Security) Act addresses AI-driven voice cloning by expressly protecting an individual’s voice (and simulation of voice) as part of their right of publicity. In addition, the law makes it illegal for a person or company to distribute or publicly dis -
341 CHAMBERS.COM
Powered by FlippingBook