Skip to content

An employee shares patient data with a public chatbot against the rules. That is a data breach.

The Dutch data protection authority received a report of such an incident at a GP practice. With AI that runs in your own environment, data does not leave the company.

The problem

An employee pastes a customer email, a patient record or a quotation into a public AI chatbot to get something done quickly. That puts personal data in the hands of an outside party, often outside Europe, where it may be stored or used to train the model. The Autoriteit Persoonsgegevens (the Dutch data protection authority) has already received several reports of such data breaches, including from a GP practice and a telecoms company, and is calling on organisations to make clear agreements.

Why it keeps happening

  • AI chatbots are free, easy and genuinely useful.
  • There is no approved alternative, so people use what is available.
  • Not everyone realises that typing into a chatbot is the same as sharing data with a third party.

What the AI Act requires

Organisations that use AI must promote the AI literacy of the people who work with it. Tailor the explanation of what it can do and what the risks are to their work. That applies to SMEs too.

Our approach

  1. Local AI as the approved alternative. An AI service that runs on your own hardware or in your own environment. Data does not leave, and there are no costs per question.
  2. Clear agreements. Which data may go into which AI application, and which may not.
  3. Permissions per application. The AI only gets access to what it needs, and may read and suggest, but not act on its own.
  4. Explanation for employees. Short and practical, so the safe route is also the obvious one.

What to watch out for

  • A ban without an alternative does not work; people will use AI anyway. That is called shadow AI.
  • Record which AI applications exist and what they are for.
  • If it happens anyway: you may be required to report a data breach involving personal data.

Would you like advice on working safely with AI? Request a free advice conversation.

Sources

Updated on

AI

Personal data in the AI cloud: who is watching?

A conversation with an AI service in the cloud is not private by default. A case in Canada shows what a provider can see and share. What does that mean under the GDPR, in the United Kingdom and in Switzerland?