With the rise of AI, abusing personal data has never been easier. Yet, parents still share their kids' data on social media. To change their sharing behavior, we created the ultimate child influencer: a grown-up version of 9-year-old “Ella” – with the same AI and deepfake technology used by scammers.

We confronted them with a cautionary tale, functioning as a sharable centerpiece: directly addressing parents on the platforms they use for sharing, spreading into organic messaging channels like WhatsApp. And providing educational tools like a social filter to pixelate children’s faces.

Social media exists to share one’s life with others, and there is nothing more important in a parent’s life than their children. By the age of five, the average child has 1,500 pictures of them on social media – shared without their consent. This is called “sharenting”.

With the current rise of AI making it easier than ever to abuse this data, The New York Times warns that by 2030, two-thirds of identity fraud cases will be connected to sharenting.

Even though lawmakers around the world are slowly turning their attention to this matter, the mass audience is still mostly unaware of the dangers of carelessly sharing children’s pictures and videos online. 2 out of 3 parents have followers and social media connections they’ve never met in real life before – making it easy for criminals to infiltrate their private circles.

2023 was the year AI tools became easy to use for everyone, raising a lot of concerns about the safety of the data we share in social media.

Yet, parents still share their kids' data on social media, unaware of potential consequences on their future: identity theft, scamming, harassment, deepfake pornography and child sexual abuse material.

As Europe’s leading telco brand, Deutsche Telekom’s ambition is to connect everyone to the opportunities of the digital world. But with the privacy of a whole generation at risk, Deutsche Telekom wanted to build on its legacy and step up to protect children against the recent arising threats.

The brief: Develop a PR-led campaign that brings awareness to the digital threat of online sharing, inspires behavioral change, and strengthens the brand’s perception as a protective force for the digital society.

Using the latest Al technology on a single photograph, we created the virtually aged version of a 9-year-old girl: Ella. In our film, she confronts her parents – and by extension parents worldwide – with the consequences of sharing kids' data online, sending a gut-wrenching warning from the future.

This way, our film became both a literal demonstration of AI manipulation capabilities and an emotional plea for the online safety of all children. As part of the campaign, we also created a solution to the problem: a social media filter pixelating children’s faces, empowering parents to #ShareWithCare and protect their children’s privacy.

The Method:

Usually, deepfake technology requires a large data set of source videos and a complex face-mapping process. But, since we wanted to stress the harm that can be done with just a single picture, our challenge was to create a whole deepfake video from a single JPEG file: a photo of 9-year-old Ella, just like those that can easily be found on social media.

We fed that photo to an AI engine that created a realistic version of a grown-up Ella by ageing her facial features. Then, we trained a machine-learning engine to recognize faces in moving images and replaced it with Ella’s aged face.

Finally, we recorded Ella’s message using a stand-in actress and replaced her face using our newly trained engine. The engine was able to replace the face and maintain all the original facial expressions and emotions.

The implementation:

We shared Ella’s heartfelt message on all relevant social media platforms – targeting parents right where they are sharing their kids’ data. Within hours, it organically spread everywhere.

From TikTok to Instagram, Facebook and LinkedIn, thought leaders endorsed our message.

Parenting influencers spread it among their communities. We made the jump to messengers like WhatsApp, turning our target audience into organic influencers themselves: advising families and friends to #ShareWithCare.

A campaign website provided parents, news outlets and institutions with educational resources for safer sharing. As a solution, we published a social media filter to automatically pixelate children’s faces when videos and pictures are taken.

Brand Relevance:

62% of non-customer parents affirmed that Telekom’s data safety and virtual privacy commitment will make them consider the company as a telecommunications provider in the future.

Ella’s message is now an educational tool used by European police agencies, such as the Scotland police and the state polices across Germany.