Learning Objectives
- Understand what data AI tools collect and how they use it
- Know which privacy settings to change on major AI platforms
- Build habits that protect your personal information when using AI
What AI Tools Know About You
When you use an AI chatbot, you are not just sending a question into the void. Here is what typically happens to your data:
What is collected:
- Every message you type and every response you receive
- Your account information (email, name, payment details if subscribed)
- Your device information, IP address, and approximate location
- How long you spend on each conversation
- Files, images, or documents you upload
How it may be used:
- To improve the AI model (your conversations may become training data)
- To personalize your experience (remembering preferences, conversation history)
- To monitor for policy violations (safety filtering)
- To generate aggregate usage analytics
⚠️Warning
The training data question. By default, some AI platforms use your conversations to train future models. This means your private thoughts, business strategies, or personal stories could influence how the AI responds to other people — not word for word, but as part of its learned patterns. This is the single most important privacy setting to understand.
Privacy Settings You Should Change Today
ChatGPT (OpenAI)
- Go to Settings (gear icon) → Data Controls
- Toggle "Improve the model for everyone" to OFF — this prevents your conversations from being used as training data
- Review Memory settings — ChatGPT remembers things about you across conversations. You can clear this or disable it entirely
- Consider using Temporary Chat mode for sensitive conversations — these are not saved or used for training
Claude (Anthropic)
- Anthropic does not use free-tier conversations for training by default — this is a key differentiator
- Review your conversation history — you can delete individual conversations or clear all history
- Pro/Team plans have additional data handling guarantees in their terms of service
Google Gemini
- Go to Gemini Apps Activity in your Google Account settings
- You can turn off Gemini Apps Activity to prevent conversations from being saved and reviewed
- Note: this is separate from your Google Web & App Activity setting
- Review saved conversations and delete ones containing sensitive information
Microsoft Copilot
- Review privacy settings in your Microsoft Account → Privacy dashboard
- Enterprise users: your admin controls data handling policies — ask your IT team
- Personal users: review what data Microsoft collects under Activity history
✅Tip
Do this now. Open each AI tool you use in a separate tab and check these settings. It takes about 5 minutes total. This is the single highest-impact action in this entire playbook.
The Privacy Spectrum
Not all AI interactions carry the same privacy risk. Here is a framework for thinking about what to share:
| Risk Level | What to Share | Examples |
|---|---|---|
| Low risk | General questions, public information | "Explain quantum computing" / "Write a poem about rain" |
| Medium risk | Work-related but non-confidential tasks | "Draft a meeting agenda for our team standup" / "Summarize this public article" |
| High risk | Personal or business-sensitive information | Financial details, health information, proprietary strategies, personal conflicts |
| Do not share | Credentials, private data of others, classified info | Passwords, Social Security numbers, other people's personal information, trade secrets |
⚠️Warning
Other people's data is not yours to share. Never paste someone else's personal information — their emails, health records, performance reviews, or contact details — into an AI tool without their explicit consent. This is both an ethical and often a legal issue.
Building Privacy Habits
Before You Type, Pause
Develop the habit of asking yourself: "Would I be comfortable if this conversation became public?" If the answer is no, either rephrase to remove sensitive details or use a privacy-focused setting (like ChatGPT's Temporary Chat).
Anonymize When Possible
When asking AI about personal situations, remove identifying details:
- Instead of: "My employee John Smith in the Dallas office has been underperforming..."
- Try: "An employee in a mid-size company has been underperforming..."
You get the same quality advice without exposing anyone's identity.
Use Paid Tiers for Sensitive Work
Most AI platforms offer stronger privacy guarantees on paid plans. If you regularly use AI for business-sensitive work, the subscription cost is worth it for the data handling protections alone.
Regularly Clear Your History
Set a monthly reminder to review and clear your AI conversation history. Even with training data opt-outs, stored conversations are a potential exposure point if the platform experiences a data breach.
Your Rights (Quick Overview)
Your privacy rights depend on where you live:
- EU/UK (GDPR): You have the right to access, delete, and restrict processing of your personal data. AI companies operating in Europe must comply.
- California (CCPA/CPRA): Similar rights for California residents — you can request deletion and opt out of data sales.
- Other US states: Privacy laws vary. Check your state's current regulations.
- Everywhere: You can always opt out of training data usage through platform settings, regardless of jurisdiction.
Key Takeaways
- AI tools collect your conversations, device info, and usage patterns by default
- Turn off training data sharing on every AI platform you use — do it today
- Use the privacy spectrum to evaluate what is safe to share
- Anonymize personal details and never share other people's private information
- Paid tiers generally offer stronger privacy protections for sensitive work
- Regularly review and clear your conversation history
