From 24 August 2026, adverts appear in ChatGPT for users in 31 European countries, including the Netherlands. That sounds like a minor change to a chat interface, but there are practical privacy considerations for organisations that use ChatGPT professionally, or do so via a free account.
What changes
OpenAI shows adverts on the Free and Go plans (€8/month). Paid plans, namely Plus (€23/month), Pro, Enterprise, Business and Education, remain advert-free. The adverts appear below the model's answer, are labelled 'sponsored' and are visually separated from the model response, according to Trending Topics.
OpenAI is rolling this out as OpenAI Ireland Limited, the European data controller.
How OpenAI tries to ensure GDPR compliance
The GDPR approach works in two layers:
Phase 1 (now): Ads are contextual, not personalised. Targeting is based on the current conversation topic, device type, location, time of day and language, without using previous conversations or saved memories. OpenAI relies on legitimate interest as the GDPR legal basis for this form. Explicit consent is not requested.
Phase 2 (later): Personalised ads based on chat history and previous ad interactions will only be activated after the user gives explicit opt-in.
Sensitive topics, health, mental health, politics, are excluded from ad targeting. Ads also do not appear in temporary chats, the browser extension or accounts identified as belonging to minors, iAmExpat NL reports.
What this means for business use
Most organisations that use ChatGPT professionally do so via a paid plan or the API, and therefore do not see any ads. But there are two practical considerations:
Employees with free accounts. If employees enter business information via personal free ChatGPT accounts, there is now an additional layer: conversation topics are used for ad targeting. In itself, that is not a GDPR violation for OpenAI, but it makes clear that a free chatbot account is not a neutral environment. Organisations that want to prevent this would do well to enforce policy on which tools employees may use for work, and via which plan.
Transparency towards your own users. If your organisation offers a ChatGPT integration to customers or employees via the API, this has no direct impact. But if you embed the ChatGPT web app or refer users to free accounts, you must inform users about the ads as part of your AI transparency obligation under AI Act Article 50, which applies from 2 August.
For GDPR-defensible use of generative AI, paid plans with a data processing agreement are the route. See our safety comparison for an overview of which models and plans offer an EU-compliant data processing agreement.
Broader picture: more models, more choice
At the same time, Grok 4.6 appeared on AWS Bedrock this week, with a context window of 500,000 tokens and configurable reasoning effort (low/medium/high). Price: $2.20/$6.60 per million tokens, according to AI Weekly. That places Grok 4.6 in the same price segment as Claude Sonnet 5 ($2/$10), but via a trusted cloud infrastructure for organisations that already run on AWS.
For Dutch organisations in regulated sectors, such as healthcare, legal and financial, the choice of infrastructure (EU region, data processing agreement) is at least as important as model quality. The model overview offers a direct comparison on price, quality and privacy profile.
Summary
Using ChatGPT for free in the Netherlands becomes ad-supported from 24 August. Business use via paid plans or the API does not change. The development makes the importance of clear AI policy in organisations concrete once again: which tools are permitted, via which plan, and what may be entered into them?
Sources: Trending Topics, ChatGPT ads Europe · iAmExpat NL, OpenAI ads Netherlands · AI Weekly, Grok 4.6 on AWS Bedrock and Cerebras CS-4 · Techtimes, opt-out ads ChatGPT Europe