ByteDance’s TikTok is reportedly testing a new generative AI feature that allows users to create digital avatars from their own photographs. The feature, discovered in the app’s code and now seemingly live for some users, represents a significant step by the social media giant to integrate personalized AI content creation directly into its platform.
According to reports, the tool prompts users to upload between three and ten photos of themselves. After processing, the AI generates several digital versions of the user in a variety of artistic styles. Users can then select their preferred avatars to use as profile pictures or within TikTok videos. The feature is being positioned as a new tool for self-expression, allowing creators to craft a unique digital persona that moves beyond standard filters.
This development places TikTok in direct competition with similar offerings from Meta and Snap, both of whom have heavily invested in AI-driven avatars and personalization tools. It also builds on TikTok’s broader AI strategy, which includes its “Symphony” AI suite for advertisers, demonstrating a clear intent to leverage artificial intelligence across both its consumer-facing and business products.
However, the feature immediately raises questions about data privacy and usage. It is not yet clear how ByteDance will store the uploaded user photos or the resulting AI-generated images. Concerns linger about whether this biometric data could be used to train the company’s future AI models, a point of contention for many generative AI tools. Given the intense regulatory scrutiny TikTok already faces globally regarding its data handling practices, the rollout of such a data-intensive feature will be closely watched by privacy advocates and government bodies alike.
English 
Deutsch