Artificial intelligence (AI) boomed a few years ago and slowly seeped into people’s lives so much that many adults today rely on it to complete their everyday tasks. It first entered under the guise of being an internet bot with answers or solutions, many a time arbitrary ones, and then eventually asked us if we would like to alter our images and videos either sparingly or in totality. From the viral Gemini vintage red sari trend to the more recent trend of blending a person’s childhood picture with their adult self, AI has made its presence unmistakably clear.
But along with these trends also came reels of content creators urging people to try the ‘restyle’ option on Instagram. What appears to be a simple photo editing feature is now often powered by AI, and many who consciously refrained from uploading their photos to Gemini or other AI platforms have unknowingly used such features, believing they pose no risk. In reality, these tools can still collect and store facial data from uploaded images, which may then be used to further train AI systems or improve facial recognition technologies. Rijin Reji, Full Stack Software Developer, expands on this.
Excerpts follow:
Why is using AI even if only to edit photos on Instagram a threat to individuals?
There is a real concern here because facial data is very personal. AI systems can examine facial structures, expressions, skin details, emotions, age estimation, and other biometric patterns. Whenever an AI trend gains popularity, AI improves quickly. When millions of people start using a particular AI effect or style, the systems get exposed to many more examples and patterns. Naturally, the output quality improves dramatically. The important thing is learning how to use these systems responsibly and safely.
For a layperson, can you explain what Meta states in its policy about what it does with images and videos that people enhance with AI options?
Meta is openly saying that when you use something like a restyling feature, your face is being analysed. Meta has reserved the right to use user content photos, videos, and text to train its AI systems. And data from Facebook, Instagram, and WhatsApp can now be combined more seamlessly for advertising and personalisation.They are not hiding that part. What they are hiding is the full picture. Their policy does not specify whether users get control over how their images are analysed or which features get activated. It is also unclear whether you are consenting to all potential uses — simultaneously summarising, modifying, generating new content, or whether you can opt in or out of specific applications. And critically, the policy does not address whether you can withdraw your consent after sharing an image.
It does mention that you can delete your personal data associated with AI interactions, but adds that deleting “may not delete Meta’s copy of your personal information”. The recent 2025 update did not clarify any of this either. It actually made things broader, giving Meta more permissions on their end while leaving users with the same vague, unresolved terms.
The consent is technically there but it is manufactured consent, not genuine informed choice. And when it comes to something as sensitive as your facial data, that distinction matters enormously.
Do you think that social media platforms should include an advisory for awareness?
The growth of AI over the last few years has been extremely fast, almost exponential. The adoption happened so quickly that awareness and digital education did not keep pace. If companies are heavily promoting AI features and encouraging millions of people to use them, they should also put effort into educating users responsibly. Not fear-based warnings that scare people. Straightforward communication explaining how uploaded data may be used, and giving users control over permissions and privacy settings.
How do you see AI’s integration with social media platforms and what do you think the future of this integration would look like?
I believe this integration will grow much deeper in the coming years. Right now, AI mostly helps users edit photos or improve content quality. Eventually, AI could generate entire videos, personalised avatars, virtual outfits, voiceovers, or even create digital versions of people for social media interactions. At the same time, we either improve current technology or invent something even better than AI itself. I don’t think anyone can confidently claim AI is the final stage of technology.
