A new AI-generated video ad campaign amassing millions of views on social media may make many parents reconsider what information they share publicly about their children.
The Deutsche Telekom video uses the latest artificial intelligence technology to virtually age a 9-year-old they call Ella, played by an actor, in order to illustrate how a child’s digital footprint can potentially be misused to damage their lives.
“I know for you these pictures are just memories, but for others they are data,” the AI-generated Ella tells the couple playing her parents in a staged social experiment. “And for me, maybe the beginning of a horrible future.”
Ella’s parents in the video, also played by actors, are described as often posting about her life online, sharing carefree moments unaware of the potential harm that it could cause her in the future.
While the video is fake, some experts say its cautionary message is very real.
“Once an image is shared online, it can be hard to control where it ends up,” explained Dr. Rebecca Portnoff, director of data science at Thorn, a nonprofit that fights the spread of child sexual abuse online, in an interview with ABC News.
The campaign documents the various potential dangers of oversharing your children’s data, from online bullying to identity theft to the creation of child sexual abuse material, or CSAM.
“Bad actors use a variety of content manipulation technologies and services to alter benign photos and videos, usually from a child’s social media account, or on the open web, to sexualize this content,” added Portnoff.
While these dangers have been documented in the past, the risks are exacerbated by the increasing rise and accessibility of AI. Paired with troves of publicly available data, this is giving criminals more content to work with.
“Sharenting” is a term sometimes used to refer to parents sharing content about their children online. It’s an easy and accessible way for parents to connect with each other, explained experts at Boston Children’s Digital Wellness Lab who have studied the potential impacts of sharenting.
As illustrated in the Deutsche Telekom campaign, generative AI is making it possible to create believable replicas of your child’s likeness from voice to figure.
The campaign also details how AI algorithms can automatically identify and collect images of children, and how they can then be used to create profiles of the children that can be used for advertising purposes or even create pornographic content.
Experts like Portnoff say that AI is making it easier for predators to both create realistic material from images available publicly or to create entirely new material.
At the same time, organizations like Thorn have also developed machine learning algorithms to build tools that detect, remove, and report CSAM.
“As the world of social media continues to evolve, parents can strive to strike a balance between connecting with friends and family and respecting kids as individuals who will have to deal with the consequences of a life lived online,” urged experts at Boston Children’s Digital Wellness lab.
As for how parents might strike that balance, sharing photos and content only to private albums with family and friends and not publicly is one way to share while safeguarding your child’s privacy and minimizing risks.
Portnoff also had a few suggestions about how families can educate kids about online safety.
“Parents can talk early and often to their kids about online safety and foster an open and judgment-free environment at home. They can work to understand the apps, social networks, and online services their child uses, as well as what privacy controls are available — and then utilize those privacy controls,” she said. “It’s important to educate children about online safety, including why it’s important to avoid sharing personal details and … the permanence of what’s shared online.”