Why Your Data Is AI’s Fuel
To understand the risk, we first need to understand the deal we are making. When you interact with a generative AI like ChatGPT or Google Gemini, you are providing it with data. These models learn from vast amounts of information, including conversations
with users. By default, many services use your prompts and dialogues to train and improve their systems. This means your data—your questions, your uploaded text, your ideas—can be stored on company servers, sometimes for years. While this helps make the AI smarter, it also creates a permanent record that could be reviewed by employees or, in a worst-case scenario, exposed in a data breach. The fundamental rule is to treat any public AI chatbot as a public forum, not a private confidant.
The Do Not Share List for Indian Users
If you wouldn't post it on a public website, do not enter it into an AI prompt. For users in India, this is especially critical. Never share Personally Identifiable Information (PII). This includes, but is not limited to: Aadhaar or PAN card numbers, passport details, or driver's license information. Keep your financial data offline; never paste bank account numbers, credit card details, UPI IDs, or investment statements. Your personal contact information, such as your full home address and personal phone number, should also be kept private to avoid risks of phishing or scams. Finally, sensitive health information—diagnoses, medical records, prescription details—should never be shared. AI is not a doctor, and sharing this data creates unnecessary privacy risks.
Beyond Text: The Hidden Data Risks
The information we share isn't just limited to the text we type. Modern AI tools can analyse images, voice recordings, and even documents you upload. When you ask an AI to summarise a PDF of a business report or a scanned personal document, you are sending that entire file to its servers. This can include confidential business strategies, client information, or personal legal matters that should remain private. Be especially mindful of metadata—the hidden information within your files, like your location, device details, and timestamps—which may also be collected. The convenience of getting a quick summary is not worth the risk of exposing sensitive company or personal data.
Your AI Privacy Toolkit: Key Settings to Check
The good news is that you have control. Major AI platforms allow you to manage your data privacy, but you have to be proactive. Take a few minutes to check these settings on the platforms you use most often. For ChatGPT, go into 'Settings' and find 'Data Controls'. Here, you can turn off the option that allows your conversations to be used to 'Improve the model for everyone'. For Google Gemini, you need to go to your Google Account's 'Data & Privacy' settings and manage 'Gemini Apps Activity'. You can pause or turn off this activity to prevent your chats from being saved and reviewed. Many platforms also offer a 'temporary chat' or 'incognito' mode, which does not save your conversation to your history. Use this feature for any query you wouldn’t want to be linked back to your account.
Generalise Your Queries for Safety
You can still get great value from AI without handing over your personal life. The trick is to learn how to generalise your prompts. Instead of asking, “Review my portfolio statement from HDFC Bank and tell me what to sell,” ask, “What are the pros and cons of investing in large-cap mutual funds in the current market?” Instead of uploading a legal contract and asking, “Is this clause in my rental agreement for my flat in Koramangala, Bengaluru, fair?” ask, “What are common things to look out for in a standard rental agreement clause about annual rent increases in India?” By removing the specific, sensitive details, you get the expert knowledge of the AI without compromising your personal information.
