Protect your personal data on the internet!

  Updated on  2 November 2026

Online purchases, AI searches, subscriptions to loyalty cards, responding to a sales offer or a survey… Whether in France or another country of the European Union, you are asked daily to share your personal data. However, are you sure that your data is always protected? Do you know who is responsible for the confidentiality of your data? How to protect them better? All the information on the GDPR and the protection of your personal data in Europe in our FAQ.

The word
© Pexels

Questions and Answers

What should I do if my data has been stolen from a third party website?

It’s possible that a website where you have entered personal data falls victim to hacking. Here is what to do if your data is leaked:

  • Contact the French National Committee on Data Protection: CNIL-Complaints | CNIL and cybermalveillance.gouv.fr. The website from which your data was stolen is, however, responsible for taking the necessary action.
  • Ask the website for detailed information about the incident: exact dates and context, origin of the attack, exactly what data did the hacker steal (identity, contact details, financial information, social security number, etc.). What documents did the hacker take (copies of your ID card, your payslips, your bank documents…)? Has the website from which the data was stolen from made a declaration to the CNIL?

Make sure to act quickly so that you can prepare for the possibility of identity theft from your data.

Many locks scattered
© Adobe Stock

How can you protect your data when dealing with an AI agent?

To personalise its results, artificial intelligence can use a great deal of information about you. The more information you provide to an AI agent, the more careful you need to be about how it uses and stores that data. Before entrusting any personal information to an AI, ask yourself whether it is necessary.

Check your privacy settings

Before using an AI agent, check its privacy policy. Whether you have a free or paid account, you should be able to restrict the use of your personal data via the settings.

These settings may include:

  • The use of your conversations to train or improve the AI.
  • How long your conversations are retained.
  • Whether the AI can retain information across multiple conversations.
  • Access to your calendar, photos, camera or microphone.
  • The use of your location data or browsing history.
  • Connection to third-party apps or services, such as your email or online storage.

Many other settings vary depending on the specific AI. Some of these settings are usually enabled by default for AI agents. So please take a few moments to adjust them to suit your needs.

Be careful with sensitive data!

Be wary when carrying out a search or asking AI a question. It is best to avoid, as far as possible, sharing sensitive data about yourself or your loved ones. For example: information relating to your health or your personal and financial circumstances.

Of course, never give the AI any passwords, usernames, reference numbers or other confidential information. You should also avoid sending it photos of yourself. The risk is that any leak of this data could facilitate an attempt at fraud or identity theft.

Example: if an AI service were to be hacked, and you had provided details of your bank account to an AI agent (name of your bank or latest direct debit), you might receive an email or text message seeming to come from your bank. The message, which looks very genuine, asks you to update certain details. It encourages you to click on a link to “resolve the issue”. This link actually leads to a fraudulent website designed to steal your bank login details or payment codes.

What happens to your conversations?

Some AI agents may use your conversations to improve their models and responses, regardless of whether you were logged in to make your queries or not. In some cases, staff responsible for service improvement may also review these conversations.

Although this data is supposed to be anonymised, it is not impossible for personal information to remain by mistake. If you wish to avoid this risk, try disabling the use of your queries and conversations for training purposes.

Very often, this use of your queries for training purposes is enabled by default in the settings. In certain situations, you can also object to the use of your personal data for training an AI agent.

Is deleting a conversation with an AI agent enough to delete your data?

No, not necessarily. If your conversations have been used to train the AI, what has been learnt from this data forms part of the system’s “knowledge” and, in practice, cannot be deleted in the same way as a simple conversation.

Some AI agents may build up a memory containing information about you, allowing them to personalise their responses. This memory can thus form a sort of profile of your interests, habits or preferences. It is generally more comprehensive and longer lasting when you are logged into an account. Without an account, the memory may exist in a more limited form, for the duration of a session, and be automatically deleted afterwards.

Deleting a set of conversations is not, in principle, sufficient to delete this data. Where deletion is possible, please check the following separately in your account settings:

  • Your conversation history;
  • The AI agent’s memory;
  • Other data associated with your account.

If you wish to delete your account and all associated data, please consult the privacy policy for specific instructions. In any case, the GDPR entitles you, under certain conditions, to be informed about the processing of your personal data. You may request access to your data, its rectification or erasure, or the restriction of certain processing activities.

Europe is aware of the commercial value of personal data

European legislation acknowledges that personal data has marketing value for the seller that collects them. Consumers who use free services usually pay with their personal data. Therefore, the legislation outlines that they must also have rights.

For example, the legal guarantee of conformity applies for any digital content or service, even if it is free. This means you can ask an app, software or subscription service to be fixed if there are problems, such as a streaming platform that is not performing correctly.

Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Innovation Council and Small and Medium-sized Enterprises Executive Agency (EISMEA). Neither the European Union nor the granting authority can be held responsible for them.