The Children’s Advertising Review Unit (CARU) is putting advertisers, brands, toy-makers, marketers and developers on notice that its advertising and privacy guidelines apply to the use of artificial intelligence in kid-targeted advertising and personal data collection.

The warning has been issued because AI usage is becoming increasingly commonplace in online advertising and data collection practices, according to Dona Fraser, SVP of privacy initiatives at US nonprofit BBB National Programs, which runs CARU.

“We call on marketers, brands and developers to prioritize transparency, safety and compliance with CARU’s advertising and privacy guidelines…[in order] to maintain the well-being of children as AI becomes commonplace,” says Fraser.

CARU cautions that it will strictly enforce its advertising and privacy guidelines in connection with the use of AI and any potential harm to children in terms of manipulative practices, including influencer marketing, deceptive claims and privacy practices.

The review unit is the first and longest-running Safe Harbor Program in the US under the Children’s Online Privacy Protection Act (COPPA). As a monitor of child-directed media, when CARU finds that data collection practices are misleading, inappropriate or inconsistent with laws and guidelines, it asks companies to voluntarily make changes and, if necessary, will refer transgressions to the appropriate federal regulatory body for enforcement—usually the Federal Trade Commission (FTC) or a state Attorney General.

In its warning, CARU notes that brands using AI in ads should be extra-cautious about potentially misleading or deceiving children in four areas:

1. AI-generated deep fakes; simulated elements including realistic depictions of people, places or things; and AI-powered voice-cloning techniques.

2. Depictions indicating product or performance characteristics that include copy, sound and visual presentations generated or enhanced using AI.

3. Fantasy footage created using animation and/or AI-generated imagery that could unfairly exploit a child’s imagination, establish unattainable performance expectations, or exploit a child’s difficulty in distinguishing between the real and the fanciful.

4. The creation of character avatars and simulated influencers that directly engage with children and could mislead them into believing they are engaging with a real person. 

Additionally, CARU stresses that if advertisers use generative AI to depict people, they should ensure these portrayals reflect the diversity of humanity and don’t promote harmful negative stereotypes.

In terms of online data collection and other privacy-related practices for online services that target under-13s, CARU wants brands using AI to exercise caution in a few key areas, including the gathering of personal information from children under COPPA and the reliance on third-party generative AI technology to operate and process data, which may require verifiable parental consent (VPC).

CARU also wants AI-connected toys and online services to collect VPC and properly disclose collection practices in their privacy policies prior to any collection, use or sharing of children’s personal information through their own online service or through a third-party generative AI service.

As for next steps, Fraser says CARU is working with industry supporters including Disney, Hasbro, LEGO, Spin Master, Coca-Cola and Super Awesome to develop additional guidelines for the use of AI in the kids space that will be published at a later date.

The full AI compliance warning is available at assets.bbbprograms.org 

Image courtesy of Possessed Photography/Unsplash