Protecting data and building trust

When you finish, you can decide what data is safe to share with an AI tool and describe your findings honestly.

Spreadsheets hold private things

A spreadsheet can look boring and still be sensitive. It might hold customer names and phone numbers, employee pay, student grades, patient details or business plans. Once you upload a file to an AI tool, you may be sharing it with a company you do not control.

Policies differ between tools and plans. Some tools may keep your data or use it to improve their services unless you change a setting, while business versions may offer stronger protections. Read the tool’s current help pages and your employer’s rules before you upload.

Remove, mask or switch tools

You can often get help without sharing real data. Delete columns you do not need, such as names, addresses and ID numbers. Replace names with codes like Customer 1. Share ten sample rows instead of ten thousand. Or invent fake data with the same layout.

For example, if you need a formula for a payroll sheet, ask for the formula using made-up numbers and test it yourself on the real file. If you work in health, law, education, finance or government, never put identifying information about patients, clients, students or citizens into a consumer AI tool unless your employer has approved it for that data. Follow your employer’s policy and the rules of your profession, including privacy laws such as HIPAA in the US or GDPR in Europe.

Be honest about what AI did

When you share findings, you are responsible for them. If AI helped you analyze data, check the key numbers yourself and be ready to explain how you got them. “The tool said so” is not a reason your boss or customer can rely on.

Also watch for bias. If your data leaves out some groups of people, or if one category is tiny, a summary can hide that. Ask the AI: “What could be missing or misleading about this analysis?” It will not always know, but the question helps you think.

Worked example: a school survey

A teacher in Accra has a spreadsheet of parent survey replies with names and phone numbers. She wants help summarizing them. She makes a copy, deletes the name and phone columns, and checks that free-text answers do not contain identifying details. Then she asks for themes.

She reads the themes against ten replies she picked herself, and notes in her report that responses came from only some families, so the picture may be incomplete. She protected privacy, verified the result and stated the limits. That is responsible data work.

Practice now

Open a spreadsheet you use and list which columns would be private if shared. Make a copy with those columns removed or replaced by codes.
Write two honest sentences you could add to a report that used AI help, such as how you checked the numbers and what the data might be missing.

Use any AI assistant you have. Do not paste private, client or patient information into it.

Check yourself

1. How can you get formula help without sharing real customer data?

2. A health clinic worker wants AI to summarize patient notes in a spreadsheet. What is the right approach?

3. What should you do before presenting numbers AI helped produce?

Ask the AI tutor

Ask anything about this lesson, in any language. The tutor is an AI: it can be wrong, and it never needs your personal details.

Take the final exam

AI-drafted. Checked by an automated validator and a separate AI reviewer; a person has not read every lesson. Translations and tutor answers are AI-made. Found a mistake? Tell us. AI can be wrong. General information only, not professional advice. Not affiliated with any AI company named. Current University is not accredited.