Beeldbank.nl: AI Training, Data Protection and Your Photos

When your organisation uploads thousands of photos to an image bank, a reasonable question follows: does the vendor use our images to train AI features? This article gives you a vendor checklist and then shows what Beeldbank.nl states on the subject. It is a reading aid for a privacy assessment, not a legal opinion.

Why Training on Your Photos Matters

Features such as face recognition and tag suggestions are common in image banks. Models like these are built from data, so it is fair to ask whose data was used and whose data may be used in the future. Teams that manage employee portraits, event photos or brand material want to know whether their uploads can end up as training material.

The question gets sharper when photos show people. Under the GDPR, personal data is any information relating to an identified or identifiable natural person. A photo of a recognisable face is such information, which is why your privacy officer will want an answer about training before you sign.

The Core Question for Every Vendor

Ask: "Do you use our photos or our data to train AI models or machine-learning applications?" Then ask for the answer in a document, such as the terms or the privacy statement, instead of in a sales conversation. Note whether the answer is a plain yes, a plain no, or a conditional statement, and what the condition is.

Beeldbank.nl states it does not use customer data to train AI models or machine-learning applications unless the customer has given explicit prior permission. The wording in its terms also names purposes it does allow: the execution, security, support and improvement of the services. Read that sentence in full when you review the terms, because the list of purposes is part of the commitment.

A permission-first statement is neither a promise of deletion nor a claim of legal compliance. It tells you what the vendor says it will do, and it gives you something you can point to later.

What Beeldbank.nl's Privacy Statement States

Beeldbank.nl's privacy statement confirms it does not use personal data and customer data to train AI or machine-learning models without prior permission of the customer or data subject.

When you evaluate vendors, note also which words they do not use. A sentence such as "your data is secure" or "we take privacy seriously" does not answer the training question. Insist on a specific answer. For a broader set of criteria across all functions, Business Photo Management Software with AI provides a full checklist for marketing teams.

Switching AI Features On and Off in Beeldbank.nl

Separate from the training question is the question of what runs on your uploads. In Beeldbank.nl, Tag suggesties and Gezichtsherkenning can each be switched on or off in the preferences. The defaults are set under Instellingen > Bedrijfsinformatie > Voorkeuren, and during an upload a slider at the bottom left switches face recognition off for that upload.

These switches are workflow choices, not a replacement for the no-training statement. You might switch off face recognition for a batch that contains no people, or because a team prefers to tag by hand. For a closer look at how suggestions and manual work fit together, read Tag Suggestions vs Manual Tagging.

Beeldbank.nl says its tariff includes all functionality, from face recognition and easy sharing options to AI tag suggestions and watermarks. For your assessment that is a pricing fact, and it does not answer the training question either way. Keep the two questions apart in your notes.

Controller and Processor: Who Decides What Happens to Your Data

For personal data in a customer's environment, Beeldbank.nl says the customer acts as controller and Beeldbank as processor. This division of roles is central to the way the GDPR distributes responsibility. The controller decides why and how personal data is processed, and the processor works for the controller.

The GDPR requires a controller to use only processors that provide sufficient guarantees of appropriate technical and organisational measures. In practice that means you should ask for the guarantees in writing, and have your privacy officer judge whether they are sufficient for your photos. This commitment from Beeldbank does not settle the legal assessment, but it gives you something specific to evaluate. For questions about searching and filtering those photos once approved, see Advanced Search in an Image Bank.

Vendor Comparison: Privacy and AI Features

What to Check	Question for Any Vendor	Beeldbank.nl's Answer
AI model training	Do you train AI models on our photos or data?	Not on customer data unless the customer gave explicit prior permission
Machine learning	Is any data used for other machine-learning applications?	Same statement covers machine-learning applications
Roles	Who is controller and who is processor?	Customer is controller, Beeldbank is processor
Feature switches	Can we switch AI features off, by default and per upload?	Tag suggesties and Gezichtsherkenning can be switched on or off; slider switches face recognition off per upload
Cost of AI features	Are they included or sold separately?	All functionality is included in the tariff

Building a Privacy-Aware Photo Workflow

Before evaluating vendors, write down which photos in your library contain people and which do not. That tells you how much weight the training question carries for your organisation. A library of product shots needs a different conversation than one full of employee and client portraits. Also decide who on your side signs off on vendor commitments, so the responsibility is clear.

Your own workflow is as important as the vendor's statements. How do you record consent from people in your photos? Who decides which images may be shared outside the organisation? Who is allowed to switch AI features on or off? None of this is solved by a vendor commitment, and all of it matters in an audit. For guidance on how AI tagging works in practice and what happens when you skip the automatic tags, read Automatic Tagging for a Photo Archive.

Making Your Vendor Choice

Data handling is for many teams a deciding factor next to features and price. Put the training question to every vendor on your shortlist, in the same words, and compare the answers side by side. Beeldbank.nl's permission-first approach is a clear competitive advantage on privacy. A vague or missing answer from another vendor tells you something too. Compare what you have in writing and make your choice based on the facts.