[REVISION HISTORY]

Privacy risks in viral AI photo trends

Updated 2 times since CLSTR started tracking revisions of this situation.

What changed

2026-09-19 22:23 UTC → 2026-09-22 08:25 UTC · added ~~removed~~

Authorities and cybersecurity experts have raised alarms regarding viral social media trends, such as the “#80lerchallenge,” which use artificial intelligence to transform user selfies into retro-style portraits. Turkey’s Personal Data Protection Authority (KVKK) warned that these AI tools may process facial images as biometric data. Concerns include a lack of transparency regarding how third-party applications store, analyze, and share high-quality facial features, such as proportions and expressions. The KVKK also noted risks of phishing attempts using fraudulent emails that mimic services like ChatGPT. In a similar warning, Kaspersky highlighted that the high-resolution images required for these trends can contain enough data to extract biometric characteristics. Because facial features cannot be changed if compromised, the potential for long-term security risks is significant. Additionally, experts noted that selfies may inadvertently reveal sensitive background information, such as documents, addresses, or credentials, which could facilitate identity theft or fraud. Recent ~~assessments by Kaspersky reiterate~~ concerns have expanded to include the privacy of minors. Germany’s Federal Criminal Police Office (BKA) and privacy advocates have warned against the risks of sharing children’s photos online, noting that ~~users often upload high-resolution, clear facial~~ once images are published, it is nearly impossible to ~~achieve realistic results,~~ control who saves, copies, or modifies them. This risk is amplified by the rise of AI-generated imagery, such as ~~1980s aesthetics or ancient Greek characters. There are ongoing concerns that these images may be stored~~ stylized portraits or ~~utilized to train~~ holiday-themed caricatures. Experts note a generational divide between those viewing AI ~~systems without explicit user knowledge regarding~~ imagery as harmless entertainment and those seeing it as a significant threat to biometric data ~~retention or secondary uses.~~ and context control.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.