If you have been scrolling Instagram lately, you would have probably noticed it, your friends suddenly glowing like they have just walked straight out of a 90s Bollywood movie, retro sarees flowing, glossy filters in full effect. No, they did not raid their mother’ closets. They were playing with Nano Banana, a viral AI filter that transforms everyday selfies into dreamy 3D avatars.
On the surface, it felt like harmless fun. Snap a selfie, type in a playful prompt, ‘Bollywood heroine,’ ‘K-pop idol,’ or even ‘anime lead’, and within seconds you have a slick, airbrushed version of yourself. Magical, uncanny, and addictive but there is more that you may not be aware of.
So behind those portraits sat Google’s Gemini model. The same AI system known for drafting essays, debugging code, or analysing scientific papers was now powering a filter that turned Instagram into a playground of fantasy avatars. And to make sure the fantasy does not get confused with reality, every image created with Gemini carries an invisible SynthID watermark and metadata tags. With the latest Gemini 2.5 Flash Image model, those digital watermarks are embedded directly into the pixels, invisible to the eye but detectable when verified, so each portrait can be clearly identified as AI-generated.
Yet as fun as these transformations seemed, sometimes Gemini’s uncanny attention to detail went a little too far.
The Mole That Sparked Questions
It started with creator Jhalak Bhawnani, who decided to test the trend. She uploaded a selfie in a full-sleeve suit and asked Nano Banana to reimagine her in a retro saree look. What she got back was picture-perfect, like a poster from a classic Bollywood film.
So far, so expected. But then viewers spotted something strange: the AI-generated image showed a mole on her arm, in the exact spot where she has one in real life. The catch? In her original selfie, her arms were completely covered.
Bhawnani called the result “scary and creepy,” asking aloud what many were already thinking: “How did Gemini know?” Her video went viral, sparking speculation. Was this AI model simply lucky at guessing, or was it pulling from something deeper?
View this post on Instagram
Where the Magic Starts to Feel Risky
The saree mole incident became less about one glitch and more about what it revealed: how much trust we place in AI without even realising it.
Every time you upload a selfie, you’re sharing more than a face. Behind the scenes, those images and prompts can carry metadata or traces that don’t just vanish into thin air. Gemini might be processing them on your device, but the fine print often includes data retention, logging, or optional sharing to improve the model. Most of us never bother to check.
And when a trend explodes, opportunists can move fast. Within days, knockoff ‘Banana-style’ apps could surface, shady downloads that might demand payments, OTPs, or permissions that have little to do with filters and everything to do with stealing personal info.
As per reports, Police in Punjab even issued warnings, reminding people how easy it is for scammers to use selfies for fake profiles, biometric misuse, or worse, financial fraud.
IPS officer VC Sajjanar put it bluntly: “With just one click, the money in your bank accounts can end up in the hands of criminals.”
ఇంటర్నెట్లో ట్రెండింగ్ లతో జాగ్రత్త!
‘నానో బనానా’ ట్రెండింగ్ మాయలో పడి.. వ్యక్తిగత సమాచారాన్ని ఇంటర్నెట్ లో కుమ్మరిస్తే ఇలాంటి మోసాలే జరుగుతాయి.
ఒక్క క్లిక్తోనే బ్యాంక్ ఖాతాల్లో ఉన్న డబ్బు నేరగాళ్ల చేతుల్లోకి పోతుంది.
ఫేక్ వెబ్సైట్లు, అనధికార యాప్లకు ఫోటోలు/వ్యక్తిగత… pic.twitter.com/ZqmsqZ0wOi
— V.C. Sajjanar, IPS (@SajjanarVC) September 14, 2025
Then there’s the bigger shadow hanging over it all: deepfakes. The more realistic these AI-generated portraits become, the easier they are to weaponise, for fake IDs, scam accounts, or manipulated images designed to harass. A tool like Nano Banana may be branded as playful, but the same realism that makes it fun is what makes it dangerous when misused.
More Than Just a Filter
In the end, Nano Banana was never just about sarees or selfies. It was a glimpse into how AI, once locked away in labs and tech demos, is now woven into everyday culture, shaping how we look, share, and even how we’re seen.
For users, it feels simple: a few taps, a shiny new avatar. But behind that simplicity is an advanced model like Gemini, one that quietly learns from and reimagines the rawest parts of our identities, our faces, our styles, even the details we never intended to show.
Maybe that’s the lesson of the Nano Banana craze. AI is most enchanting when it feels like play, and most risky when we forget it isn’t a toy. The saree trend will fade, replaced by the next viral look. But Gemini, and the questions it raises about privacy, identity, and trust, is here to stay.
So the next time you’re tempted to hit Generate AI Portrait, remember: it’s not just a selfie you’re uploading. It’s a piece of you, being rewritten by a machine that might just know more than you expect.
