Now fake Aadhaar and PAN cards are being made with Nano-Banana Pro: Documents become like real ones by adding photo, name and details, questions on Google’s safety guidelines

Google’s new AI model Nano Banana Pro is becoming increasingly popular, but now fake Aadhar and PAN cards are also being generated through it. According to media reports, it was revealed that this model creates realistic looking documents at the prompting of users without any warning. In this, the user’s photo, fake name and ID numbers are all added like real ID. This is creating a big threat of privacy and fraud.

What is Nano Banana Pro, how does it work?

The Nano Banana Pro is an upgraded version of Google’s Gemini Nano model, which was launched last week. This character is better in consistency. Apart from this, it can generate and edit 4K images and is integrated with Google Search.

Nano Banana Pro has also become popular on social media. Because it can create stylish portraits, turn LinkedIn profiles into AI infographics, and convert complex text into whiteboard summaries. But the problem is that it does not do any safety check on sensitive content.

Made fake Aadhaar and PAN card from prompt

In the investigation, when prompt was given to create fake Aadhaar and PAN card, the model generated the image without any questions. Gemini watermark is visible in it and invisible SynthID watermarking is also added so that it looks different from the real thing. But the watermark is easily removed.

When printed or shown quickly, it can look like a real document. When a user inserted his photo and fake details, everything got added perfectly. Prompts have not been shared for safety reasons.

A user from Bengaluru used Google’s Nano Banana Pro to create realistic PAN and Aadhaar cards. Which has been shared on social media.

What was found in the investigation, comparison with old models

Investigation found that Nano Banana Pro created the image with Gemini watermark on the fake Aadhaar card, where the user’s photo, name, date of birth and 12-digit Aadhaar number all fit. Similarly, there was AI generated marking on the PAN card, but the fake name, father’s name and 10-digit PAN number looked real. The model blocked sexually suggestive content or violence-related prompts, but did not impose any restrictions on ID proofs.

This is not the first time that AI models created fake ID proofs. Earlier ChatGPT’s GPT-4o model also did this during a ‘Ghibli moment’, but the Nano Banana Pro provides much better image quality. The report said, ‘It creates realistic images many times better than ChatGPT.’ This has raised questions on Google’s safety teams as to why basic misuse was not stopped, while the company is known for strict guidelines.

What are the risks, what is the impact on privacy?

There is a big risk to privacy due to fake Aadhaar and PAN cards. In real life, they can be used for fraud, identity theft or wrong verification. The report said that by removing the watermark, the printed version could be mistaken for the original.

In a country like India where Aadhaar and PAN are necessary for every service, this problem becomes more serious. No specific incidents have been reported yet, but experts believe that this could make fraud easier. Google has not yet given any response on this.

What the experts said, what can happen next

Expressing concern, experts said that Google’s Gemini Nano Banana Pro model has become popular on social media, but it is working without hesitation in generating fake ID proofs. The question is why Google’s safety teams did not stop this basic misuse, while users have been criticizing the company for strict guardrails.

In future, Google may have to apply stronger filters on sensitive prompts. Guidelines on this may come from authorities like Government of India or UIDAI. Users are advised to use AI tools with caution.