Be Cautious of the Trend of Editing Children’s Faces with AI, IPB University Expert Warns of These Risks
The growing trend of editing children’s faces using artificial intelligence (AI) for dance-related social media content needs to be approached critically by the public.
Chairperson of the Artificial Intelligence Study Program at IPB University, Prof Yeni Herdiyani, emphasized that behind the cute and visually appealing results, there are serious risks related to children’s data security and privacy.
Prof Yeni explained that much of today’s AI technology falls under generative AI (Gen-AI), which is capable of producing data in the form of images, videos, text, and even audio based on existing models and input information.
“Babies and dance are models that algorithms can already generate. When provided with real photos, AI can automatically produce realistic looking outputs. That is how technology works,” she explained.
From a risk perspective, the use of children’s facial data in AI systems is highly vulnerable to digital misuse. Prof Yeni noted that once photos are uploaded to public spaces, the data becomes accessible and may be misused by irresponsible parties.
“This situation places children among the most vulnerable groups within the digital ecosystem,” she stated.
Therefore, Prof Yeni continued, parents need to understand the digital risks before uploading or editing their children’s faces using AI technologies.
These risks include data security, privacy protection, and the potential misuse of personal data. In addition, users should be aware that data provided to AI algorithms can be used to train models as datasets and may reappear in other forms.
In response to this trend, Prof Yeni stressed the importance of digital literacy in society. She explained that IPB University consistently incorporates ethics and responsible technology use into computational thinking courses for first year students.
These efforts are expected to foster a more critical and responsible society in utilizing AI technologies. (dr) (IAAS/CAA)
