Is it safe to share project data with AI tools?

There's no single yes or no answer. When you use a cloud AI tool, what you type or upload is sent to the provider for processing. How it is stored, how long it is kept, and whether it can be used to improve models depends on the provider and the type of account. Terms also change over time.

This article is general guidance, not legal or compliance advice.

Check these first

1. Your organisation's AI policy. Many organisations list approved tools and data types that must never go into them. If there's no policy, ask your IT or security team before sharing work data.
2. The provider's data-use settings. Read the current privacy and data-use terms for the exact account you're using. Personal and business accounts are often treated differently, and there may be settings you can change.
3. Contracts and NDAs. Client agreements may restrict where their information can go.

Classify before you share

• Low risk: generic methodology questions and templates with no identifying details.
• Medium risk: internal status, timelines, and anonymised challenges.
• High risk: client names, personal data, budgets, strategy, and anything under NDA. Only share this through tools your organisation has approved for it.

Ways to reduce what you share

• Anonymise. Use "Project Alpha" and role titles instead of client and people's names.
• Share the minimum. A scheduling question doesn't need your full charter.
• Summarise first. Share an abstracted version instead of the raw document.
• Ask for frameworks. Ask how to approach a problem rather than pasting in sensitive data for analysis.

Chapter 11 of the book covers setting up a local AI environment with VS Code and Claude Code. Working locally gives you more control over your files, but check what each tool still sends to its provider.

For more, read Data Privacy Concerns With AI: What PMs Need to Know and The Ethics of AI-Assisted Project Management Decisions.