Privacy considerations across different tools

What data do AI tools collect?

When you use AI tools, they collect various types of information:

Account information: Email address, name, phone number (if required for verification)

Usage data: How often you use the tool, what features you use, when you use it

Conversation content: The questions you ask, prompts you provide, and responses the AI generates

Documents and files: Any documents you upload, images you generate, or recordings you transcribe

Different tools have different privacy policies and data practices. Always review the privacy policy before using a tool for sensitive tasks.

General privacy principles

Don't share sensitive personal information: Avoid including passwords, social security numbers, credit card details, medical records, or confidential business information in AI conversations.

Assume your data is processed on servers: Most AI tools send your data to cloud servers for processing. This means your information leaves your device.

Check data retention policies: Some tools keep your conversations indefinitely. Others delete data after a certain period. Review each tool's policy.

Review who can access your data: Some tools use your conversations to train AI models. Others keep your data private. Check the terms of service.

Privacy by tool type

Chatbots (ChatGPT, Claude, Gemini, Copilot):

• Your conversations are typically stored and may be used to improve AI models
• Most tools let you opt out of data sharing in settings
• You can usually delete individual conversations or your entire history
• Business/Enterprise plans often have stricter privacy guarantees

Privacy tip: Check account settings for options like "Do not use my data for training" or "Delete conversation history automatically."

Writing assistants (Grammarly, etc.):

• Process your text on their servers to check grammar and style
• May retain documents or writing samples
• Business plans often include stronger privacy protections

Privacy tip: Avoid using writing assistants for highly confidential documents unless you've reviewed their enterprise privacy policies.

Transcription tools (Otter.ai, etc.):

• Upload and store audio recordings and transcripts
• May use recordings to improve transcription accuracy
• Some tools offer local processing (like Google Recorder on Android) for better privacy

Privacy tip: Use local transcription tools (like Google Recorder) for sensitive meetings, or obtain consent before recording others.

Image generation tools:

• Prompts and generated images may be used to train AI models
• Some tools make generated images publicly visible by default
• Check privacy settings before generating images

Research tools (Perplexity, NotebookLM):

• Process your search queries and uploaded documents
• NotebookLM is made by Google, so your documents are processed by Google's infrastructure
• Always check data retention policies

Opting out of data sharing

Many AI tools let you control how your data is used:

ChatGPT: Settings → Data Controls → Opt out of using your data for model training

Claude: Similar opt-out options in account settings

Grammarly: Review privacy settings for data sharing preferences

Always check: Privacy settings change over time. Review them periodically.

When to avoid AI tools entirely

Don't use AI tools for:

• Confidential legal documents (use local tools or consult professionals)
• Private medical information (unless using HIPAA-compliant tools)
• Proprietary business information (unless using enterprise tools with appropriate contracts)
• Personal financial details like tax returns or bank statements
• Content that would violate NDAs or confidentiality agreements

For sensitive work, use tools that process data locally on your device or seek enterprise contracts with strong privacy guarantees.

Privacy-focused alternatives

If privacy is a top concern, consider tools with better privacy practices:

Local processing tools:

• Google Recorder (Android) - Transcribes offline, no cloud upload
• Local installations of Stable Diffusion - Generate images on your own computer
• Hemingway Editor (web) - Doesn't send data to servers

Tools with strong privacy commitments:

• NotebookLM (Google) and Perplexity both have clear data policies
• Some coding assistants (like Tabnine) offer local deployment for enterprises

Managing privacy across multiple tools

Keep a list: Note which tools you've given access to sensitive information

Regularly review privacy settings: Check every few months to see if policies have changed

Delete old data: Periodically delete conversation histories, uploaded documents, or generated content you no longer need

Use different accounts for different contexts: Consider separate accounts for personal vs professional AI tool usage (if allowed by the tool's terms)

Reading privacy policies

Privacy policies are long and dense, but look for answers to these questions:

How long is my data stored? (Days, months, years, indefinitely?)

Who can access my data? (Only me, the company's staff, third parties?)

Is my data used to train AI models? (Can I opt out?)

Can I delete my data? (Can I delete specific conversations or my entire account?)

Where is data processed? (Which countries or regions?)

Enterprise and team plans

If you use AI tools for work, check whether your organization has enterprise agreements:

Enterprise plans often include:

• Stricter data privacy guarantees
• Data retention controls
• No use of your data for training AI models
• Compliance with regulations (GDPR, HIPAA, etc.)

Ask your IT department whether approved AI tools exist with appropriate privacy protections.

Children and student privacy

If you're a parent or educator:

Check age requirements: Most AI tools require users to be 13+ or 18+

Review student data policies: Some tools offer special student accounts with enhanced privacy (like GitHub Copilot for students)

Supervise usage: Help younger users understand privacy risks and appropriate information sharing

Related resources

Want to know which tools to sign up for? See Do I need accounts with multiple AI tools?. Looking for organization tips? Check How to organize your AI tool usage. For getting started guidance, visit Which AI tool should I try first?.