A Swiggy Instamart customer allegedly used Google’s Gemini powered imaging tool Nano Banana Pro to turn a photo of a tray with only one cracked egg into an image showing over 20 cracked eggs.
The customer then allegedly secured a full refund of about ₹245, triggering fresh concern over AI driven refund fraud in India’s quick commerce sector.
The case, first highlighted in a viral X post by user @kapilansh, has reignited debate on how easily trust based, photo led grievance systems can be gamed when powerful image tools are freely available to everyday users.
While Swiggy has not issued a detailed public statement on this specific order, the incident has prompted social media users and industry voices to call for stronger verification, AI detection and ethical use norms so that genuine customers are not punished for the misconduct of a few.
How an edited egg photo led to a refund
According to reports and posts shared on X, the customer had ordered a tray of eggs via Swiggy Instamart and found that only one egg was cracked on delivery.
Instead of filing a routine complaint, they allegedly opened Google’s Nano Banana Pro or Gemini Nano, entered the prompt “apply more cracks,” and generated a hyper realistic image in which more than 20 eggs appeared damaged.
Screenshots shared online show the edited image being submitted as “proof” to Instamart’s chat support, which then approved a full refund of around ₹245 without questioning the authenticity of the photograph.
Commentators like product and tech professionals have warned that if even a small percentage of users adopt such tactics, quick commerce platforms working on thin margins could see their refund costs rise sharply, putting pressure on unit economics.
Social media reaction
The viral thread and related coverage sparked a mix of amusement and alarm, with some users jokingly calling it a “free money glitch” and others condemning it as an example of a “low trust society” scamming companies that try to be customer friendly.
Many commenters suggested practical safeguards, such as open box deliveries recorded on cameras, better training for support teams and integration of AI based tools that can detect manipulated images.
Several posts pointed out that Google’s AI images may carry hidden markers like SynthID and sometimes even visible Gemini branding, and urged e commerce and quick commerce firms to plug such signals into their verification systems to tell real photos from edited ones.
Although there is no widely reported official statement from Swiggy or Google specifically on this transaction yet, the debate has already grown into a wider discussion on refund fraud, responsible AI use and the need for faster policy responses.
What is Google’s Nano Banana Pro?
Nano Banana Pro is Google’s advanced AI image generation and photo editing model, built on the Gemini 3 system to create highly realistic visuals from simple text prompts.
Users can upload a photo and ask the tool to change specific details such as adding cracks to an egg, shifting lighting from day to night or altering camera angles, with the edits blended so smoothly that the final image can look convincingly real.
The Logical Indian’s Perspective
This episode is a reminder that technology itself is not the enemy; it is how people choose to use or misuse it that decides whether it builds or breaks trust in society.
When individuals exploit AI tools to fake damage and claim refunds, they do not only hurt companies, they also make life harder for delivery workers and honest customers whose genuine complaints may face more suspicion.
