GDPR Questions to Ask Before Your Team Uses an AI Tool

When an AI tool can access, receive or generate personal data, privacy questions need to be part of the decision before the tool becomes routine.

This does not mean every AI tool should be rejected. It means the organisation should understand enough about the intended use and the supplier to make a responsible choice.

The following questions are a practical starting point for a team considering a new AI tool or feature.

What personal data could be involved?

Think beyond the most obvious examples. A prompt, document, transcript, support ticket or exported spreadsheet can include names, contact details, account information, performance information, location data, special-category information or other data that relates to identifiable people.

Ask what staff will enter, what the system can access automatically and what it may generate in return.

Why are we using the tool?

Be clear about the purpose. The reason for processing information should be connected to a real business or organisational need, not simply to the availability of a new feature.

A defined purpose also makes it easier to decide what information is necessary and what information should stay out of the tool.

What does the supplier say about data handling?

Look for clear, accessible information about where data is processed, whether it is retained, whether it may be used to improve or train models, who else may process it, what security measures are in place and what controls are available to customers.

If the documentation does not answer practical questions, ask the supplier. If the supplier cannot provide enough information for your intended use, that is useful information in itself.

Can we support people’s privacy rights?

Think about what would happen if someone asks for access to, correction of or deletion of their personal data. The details will depend on the tool and the use case, but the organisation should understand where data is and how it can be managed.

Are there higher risks that need a closer look?

A new or high-impact use of personal data, a combination of data sets, use involving children or vulnerable people, or a use that may affect people significantly can require more careful assessment.

Do not assume that a supplier’s standard terms resolve your organisation’s own responsibilities. The context and intended use still matter.

What guidance do staff need?

Staff may need a short guide that says which tools are approved, what information should not be entered, when outputs need review, and who to contact with a question or concern.

Clear guidance protects people and helps the organisation use useful tools more consistently.

Take a proportionate next step

For some low-complexity uses, the next step may be to collect supplier information, set a short internal rule and train the relevant people. For more complex or higher-risk uses, the organisation may need a structured privacy and governance review, and in some circumstances qualified legal advice.

PNDAR can help organisations organise the questions, identify information gaps and plan practical next steps.

PNDAR provides practical privacy, governance and implementation support. It does not provide legal advice, legal opinions or a guarantee of compliance.