The privacy concerns around an AI app that transforms selfies into 1990s students

It is very likely that in recent days you have come across the face of an acquaintance or a famous person on social media, in the style of an American yearbook from the 1990s. All these photos have been generated with EPIK, a photo editing service which in just a few days has become the most downloaded app across mobile devices. The subject appears against a light blue background, or dressed as a cheerleader, or wearing glasses to seem more intellectual, or in a basketball uniform. These images made with artificial intelligence (AI) are all the rage on social media, where someone recently sent in a picture of Spain’s acting prime minister, Pedro Sánchez.

The application, developed by the South Korean company Snow Corporation, has experienced a total of 92.3 million lifetime installs since it was launched in August 2021, most of them in Asian countries, according to data from the market intelligence firm Apptopia. However, in recent weeks it has also become popular in the United States (4.7 million downloads) and Europe, thanks to the TikTok viral trend known as the #YearBookChallenge, in which celebrities and influencers such as Chanel have taken part.

However, the popularity of the app has reopened the debate about the safety of this type of tools, which work with AI based on the data and faces of their users. The questions follow a recent case in Spain, where a group of school girls were the victims of pornographic deepfakes made and spread by their own classmates.

The way it works is very simple. After downloading it, users have to go to the Yearbook Effect feature and upload between eight and 12 selfies so that the AI can generate photos with the desired effect. The service is not free; it costs on average $5.99, although there are price differences for standard delivery and for express service to get the photo back within two hours. Using the service also implies accepting the privacy policy, which in turn entails giving consent for the company to collect, store and process one’s facial images to provide features, content, and editing experiences.

Borja Adsuara, a lawyer specializing in digital communication, notes that the app — which works in a similar way to FaceApp, the Russian mobile program that ages faces and became popular in 2019 — presents privacy problems. The company acknowledges that it uses users’ photos to “help us develop, deliver, test and improve the application,” and that it collects information from the mobile device even when the app is not being used, but just running in the background. “The danger is not so much that other people can do horrible things with your face, as was the case with Almendralejo [the Spanish schoolgirls], but rather what the company can do with your data,” Adsuara emphasizes. “But since the result of this filter is cool, people don’t care, because they don’t see a direct danger to themselves.”

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition