Learn/AI Fundamentals/Ethics & Responsible Use
14 of 15

Privacy & Data

6 min read

Here's something most people don't think about when they paste their company's financial data into ChatGPT at midnight: where does that data go? This lesson covers what you need to know to use AI tools without accidentally giving away information you shouldn't.

The Data Flow: What Happens to Your Inputs

When you type something into an AI tool, here's what can happen to it — depending on the tool and your settings:

1

It's processed to generate your response

This is the obvious part. The AI reads your input and generates an answer. This always happens — it's the whole point.

2

It may be stored on the company's servers

Most free AI tools store your conversations. OpenAI, Google, and others keep your chat history. This means your inputs sit on someone else's servers.

3

It may be used to train future AI models

This is the big one. Many free tools include a clause that says your inputs can be used to improve their models. That means your private data could influence what the AI says to other people in the future.

Never Paste These Into Free AI Tools

Company financial data, customer personal information, trade secrets or proprietary code, medical records (yours or others'), passwords / API keys / credentials, legal documents with confidential information, employee performance data or personal details. If you wouldn't email it to a stranger, don't paste it into a free AI tool.

Free vs. Paid: The Privacy Difference

This is important: paid and enterprise AI plans typically have much stronger privacy protections:

Free Tier (Higher Risk)

  • Inputs may be used for training
  • Conversations stored indefinitely
  • Limited control over data deletion
  • No formal privacy agreements

Paid / Enterprise (Lower Risk)

  • Inputs typically NOT used for training
  • Data retention policies you can control
  • Delete conversations anytime
  • Business-grade privacy agreements

If your company uses AI tools, ask your IT department about the data policies. Many organizations now have approved AI tools with proper privacy agreements. Using the approved tool is almost always safer than pasting company data into your personal ChatGPT account.

Practical Privacy Rules

1

Anonymize before you paste

Replace real names, account numbers, and identifying details with fake ones. "Help me draft a response to John Smith about his $50K contract" becomes "Help me draft a response to a client about a medium-sized contract."

2

Use the right tool for the sensitivity level

Free ChatGPT for brainstorming and learning? Fine. Free ChatGPT for your company's unreleased product strategy? Absolutely not. Use your company's enterprise AI tool for sensitive work.

3

Check the settings

Both ChatGPT and Claude have settings where you can opt out of having your data used for training. It takes 30 seconds. Go do it now if you haven't.

Quick Check

Your colleague pastes a customer's complaint email — including their name, account number, and phone number — into free ChatGPT to draft a response. What's the risk?

Key Takeaway

Everything you type into AI tools can be used to train future models. Know what to share and what not to.