Instagram: Photographers Stop Showing Children's Faces for AI

A court ruling on AI-generated CSAM is circulating among parents and accelerating announcements from photographers who are opting not to show children's faces. Meanwhile, an incident involving AI prompts at Meta, acknowledged as a mistake and corrected, rekindles fears of misuse. Amid heightened caution and unchanged practices, the publication of children's images is undergoing a transformation.

A Renewed Ruling and Modified Policies Among Photographers

A recent ruling regarding the possession of AI-generated child pornography is circulating in parent groups on Instagram. The judge reminds that a legal precedent protects the possession—but not the production or distribution—of such images in the name of freedom of expression. This stance and its implications for photos of children shared online are widely discussed. In this context, many professional family photographers, often women or mothers, are announcing that they will no longer publish children's faces on their networks or websites. Hannah Haggard, a photographer in Kentucky, has formalized this policy, even with parental consent, explaining that the court decision has affirmed her choice, as she already refrained from publishing without consent. Attorney Mariefaye Bechrakis believes the situation is more complex, noting that judges are paving the way for an appeal and providing guidance for future prosecutors, while also reminding that current AI models are trained on vast datasets. Criminal cases involving AI-generated CSAM have already occurred, some receiving national coverage.

What Parents and Professionals Fear: Deepfakes and Face Reuse

Concerns center around the possibility that published photos of children could be used to create targeted deepfakes, or that their faces could be integrated into training datasets for AI models capable of generating CSAM.

Meta's AI Prompts on a Family Video: Acknowledged Error

Kalie Robins, a mother in Utah, drew attention to AI-generated questions from Meta under a video of her and her child, including "Who is the passenger child?" After further interactions, she noticed the reappearance of photos she thought were deleted and the retrieval of information about her children from her mother's Facebook page. Meta states that the responses from this feature are based on information already accessible to the user. Spokesperson Andy Stone acknowledges that the feature should not have posed such questions and asserts that the error has been corrected. Online, reactions deemed this behavior troubling and viewed it as a reminder to be cautious about the information and images shared by parents.

Between Heightened Vigilance and Continuity of Practices

Kalie Robins explains that she understands how Meta's AI could have aggregated this information and considers the incident a reminder of the amount of data shared, despite her own caution, such as the absence of school information or geographical markers in her posts. Other professionals, like Missouri photographer Katey Sutton, are formalizing a policy of not publishing faces and expressing doubts about the "synthetic" nature of AI, which they describe as collecting and reusing online content. However, not all parents have stopped posting photos of their children on Facebook or Instagram, even though recent concerns are prompting some to assess the risks before posting.

An Old Topic Revived by Two Recent Viral Waves

The debate over publishing children's faces on social media has long existed, but AI-related concerns are giving it a new dimension and have led Meta to intervene. Two viral events on Instagram last week prompted some parents to decide not to publish images of their children anymore. The viral video about Meta's AI suggestions contributed to this anxiety, while the fear of misuse for CSAM is encouraging parents and professionals to reevaluate their habits, even leading some photographers to adopt no-face publication policies.