What does AI mean for privacy?
With AI, both the training data and the data we type in while using it matter.
What is it?
Two flows of data matter with AI: the data a model was trained on, and the data you type in while using it. What happens to what you type depends on the provider, the type of account and your settings.
An everyday example
An employee pastes a contract containing a client’s details into a chatbot to get a summary. The data may now be stored with an outside provider.
Why does it matter?
Data you have entered are hard to take back. Health, financial and business data and information about other people are especially sensitive.
A common misconception
That deleting a conversation removes it from every system, or that local AI is always private. Check the terms and settings.
In more detail
What should I not enter without thinking?
Passwords, health data, trade secrets, children’s data or other people’s documents.
Does “delete chat” mean it is deleted from every system?
Not necessarily. Processing and storage depend on the provider’s rules, your settings, contracts and the law.
Is local AI always private?
Processing on your own device can help, but you still need to check connections, logs, the device and the security of the software.
What should an organisation do?
List its data, establish the legal basis, retention periods, access and contracts, and carry out a risk assessment where required.
Remember: Data about other people is not automatically yours to hand to a model.
Reliable sources
Last reviewed 1 November 2026. An explainer, not professional, legal or medical advice. For rules and figures that change, check the official version.