What to never tell GPT?

What You Should Never Tell GPT—And Why It Matters

As AI chatbots like GPT become part of our daily routines—helping with writing, coding, or even personal advice—it’s easy to forget they’re not human confidants. In fact, sharing too much can put your privacy and security at risk.

Never share personal identity information, such as your Social Security number, passport details, or full home address. Even seemingly harmless data, like your birthdate and mother’s maiden name, can be exploited if collected and misused. These details are often the keys to unlocking secured accounts.

Similarly, avoid disclosing specific medical results or health history. While it might be tempting to ask an AI for a second opinion on symptoms or test results, remember: these systems aren’t secure medical platforms. Your data could be stored, analyzed, or even used to train future models—without your consent.

Financial account details—like credit card numbers, online banking credentials, or even recent transaction records—should also stay private. AI models don’t need that info to help you budget or explain financial concepts, and sharing it could open the door to fraud.

Another major red flag? Proprietary company information. Whether you're drafting a confidential email or brainstorming a new product idea, remember that anything typed into a public chatbot could be retained and potentially exposed. Employees at tech firms have already accidentally leaked internal code by asking AI for help debugging.

The bottom line: treat AI like a helpful stranger—polite, informative, but not someone you’d hand your diary to. As Nguyen’s reporting and privacy experts stress, the safest data is the data you keep to yourself. When in doubt, leave it out.