What Not to Tell an AI Chatbot: A Privacy Guide

People tell chatbots things they wouldn’t tell colleagues: health worries, money problems, business plans. These tools are genuinely useful for that, but it’s worth knowing where your words go.

What happens to your chats

Depending on the service and your settings, conversations may be stored, used to improve models, or reviewed for safety. Consumer, business and API products often have different policies. Read the privacy settings of the tool you use.

Don’t share

• Passwords, OTPs, PINs or API keys.
• CNIC, passport and bank or card numbers.
• Other people’s private information without their consent.
• Confidential client or company documents, unless your employer has approved the tool.
• Unreleased financial results or trade secrets.

Settings to check

• Whether your chats are used for model training, and how to opt out.
• Chat history and memory controls, and how to delete them.
• Temporary or incognito chat modes.
• Connected apps and permissions you’ve granted.

At work

Many companies offer business versions with stronger data protections and admin controls. Use those for work tasks and follow your organisation’s AI policy.

Health and legal questions

AI can help you understand a topic and prepare questions for a professional. Share what you’re comfortable with, and don’t treat the answer as a diagnosis or legal advice.

Sources

• OpenAI, Data controls FAQ
• Anthropic, Privacy Center
• Google, Gemini Apps privacy hub