Guide

    AI that keeps your data in the EU.

    Yes, if the service processes your conversation in Europe, doesn't train on it and keeps nothing of it after the report. Why an EU data centre isn't enough for that, you'll read below.

    S

    Soren Colton · updated 5 November 2026 · 10 min read

    Short answer

    You can use AI without your data going to the US: with a service that processes your conversation in Europe, doesn't train on it and keeps nothing of it after the report. An EU data centre on its own is not enough: it tells you where your data sits, not where your conversation is processed or what's left of it afterwards.

    For confidential conversations Notuly is that kind of service: your conversation is processed in Europe, the whole AI step runs on our own models, on servers in Amsterdam, and nothing stays behind at Notuly. If it's about public texts or a meeting without names and figures, an American AI tool is often a perfectly good choice.

    Can it be done?
    Yes: processing in Europe, no training on your conversations, nothing left behind after the report
    EU data centre
    Determines where your data sits, not which law applies to the company
    CLOUD Act
    US law: American providers must hand over data, even when it sits in the EU

    It comes down to the whole chain, from the language model to the server: where each link runs, and what stays behind after the conversation. That may sound like a matter for the lawyers. Yet it's often the very first thing a data protection officer asks about, the moment an AI tool lands on the table.

    Picture an information security officer at a local authority who has to approve an AI notetaker for the executive board's meetings. She reads the privacy page, sees "data stored in a European region" and ticks the box.

    A week later the data protection officer starts asking questions. The data centre really is in Europe, but the parent company is in the United States. And that puts the entire service under American law, however European the server may be.

    To see how that happens, it helps to take an AI service apart for a moment.

    What does keeping your data out of the US actually mean?

    Keeping your data out of the US means your conversation leaves Europe at no step at all: not at the language model, not at the speech-to-text and not in storage, and that nothing of it stays with the provider afterwards. That's because an AI service isn't one thing but a stack of layers. Only once you look at them separately do you see where things go wrong.

    • The language modelThe model that writes the summary. Often from OpenAI, Microsoft or Google; sometimes the provider's own model.
    • The speech-to-textThe step that turns the conversation into words. It runs somewhere too, and it belongs to someone.
    • The cloudThe servers everything runs on, and where the data sits.
    • The companyThe owner of the service, and the country it answers to. That jurisdiction decides who can legally reach your data.

    Most AI tools have an American party on at least one of those layers: the model comes from OpenAI, the cloud belongs to Amazon or Microsoft, or the company itself is American. So ask, layer by layer, where it runs and what stays with the provider after the conversation. And ask whether your conversations are used for training.

    Sovereign AI, European AI, AI without an American cloud, AI without Big Tech: they're different words for the same question. Where is your conversation processed, and what's left of it afterwards?

    At Notuly your conversation is processed in Europe, and our own models on servers in Amsterdam create the report. Once the report has been sent, nothing stays behind at Notuly.

    What is the CLOUD Act, and does it reach EU data?

    The CLOUD Act (in full: the Clarifying Lawful Overseas Use of Data Act) is a US law that obliges American providers to hand data over to the US government, even when that data sits in Europe. That last part is the whole point.

    An American company falls under American law even when its servers are in Frankfurt or Amsterdam. The Dutch National Cyber Security Centre (NCSC) wrote in 2022 that the chance of the US actually invoking the CLOUD Act against European customers is small. But small isn't zero, and the provider can't rule it out (see the sources at the bottom).

    That became concrete on 18 June 2025, when an inquiry committee of the French Senate held a hearing. There Anton Carniaux, director of public and legal affairs at Microsoft France, was asked whether he could guarantee under oath that French customers' data would never go to the US government without consent.

    "No, I cannot guarantee that, but, again, it has never happened before."

    The quote is translated from the French, and it has indeed never happened, as Microsoft itself points out. But the provider can't guarantee it, and that's what it comes down to: the law of the home country comes first.

    So what can be demanded under that law depends on the company and its partners, and on what stays with them, not on the server. That's why the last part is what counts at Notuly: once the report has been sent, nothing stays behind, and the report sits in your own inbox.

    How to check a tool for this yourself is covered on our security page. Here we're after the why, and the why starts with a misunderstanding about the server.

    Is an EU cloud or an EU data centre enough?

    No, not automatically. An EU data centre determines where your data physically sits. It doesn't determine which law applies to the company that can reach it.

    That difference is often glossed over. An American provider puts a data centre in Europe, sticks an EU flag and a certificate on it, and calls it sovereign. In practice that's known as sovereignty-washing: it looks European, but the parent company, and with it the ultimate say, stays American.

    And the odds of running into such a set-up are high: American hyperscalers hold around 70% of the European cloud market (Synergy Research Group, 2025).

    "Our data sits in Europe" and "our data falls under European law" are therefore two different promises: the first is about a place, the second about control. So for a confidential conversation you want to know more than the place: where it's processed, whether anything is trained on it, and what stays with the provider after the report.

    At Notuly the data centre is therefore just one layer, and on top of it the whole AI step runs on our own models. We work in line with ISO standards. Our data servers run in an ISO 27001-certified data centre in Amsterdam.

    Where your conversation really goes.

    The easiest way to see this is to lay two routes side by side. Not as a checklist, but as the journey your conversation makes from the moment someone starts the recording.

    Route A: the usual route, with storage at the provider

    1. Your conversation. Someone starts the recording.
    2. Processing wherever the model runs. The tool sends the audio or the text to a model from OpenAI, Microsoft or Google, and that model doesn't always run in Europe.
    3. Storage at the provider. The recording, transcript and report stay in an account, often on an American cloud, sometimes in an EU region.
    4. Under the CLOUD Act. Whatever stays with an American provider falls under American law, wherever it sits.
    5. With free consumer AI, often: training. Your input helps improve the model.

    Route B: the route that stays in the EU

    1. Your conversation. Someone starts the recording.
    2. Processed in Europe. Our own, self-hosted models in Amsterdam create the report.
    3. The report in everyone's inbox. With the summary, the decisions and the action points, and the full transcript as an attachment, unless you switch it off.
    4. Audio deleted within a minute of processing. The report sits in your own inbox.
    5. No training. Your conversations are not used as training data, and your data never leaves the EU.

    The difference between the two routes isn't "better encryption", because both can be strongly secured. It lies in what happens to your conversation: where it's processed and what stays with the provider afterwards. Which is why an EU flag on a website proves nothing on its own.

    Notuly follows route B: the report is in everyone's inbox within ten minutes, and once it has been sent, nothing stays behind at Notuly.

    When data going to the US is perfectly fine.

    For a great deal of work an American AI tool is a perfectly good choice, and we're honest about that. Those tools are technically strong, well secured and connect to everything you already use. For a brainstorm, a public presentation or an internal note without sensitive content, where the model comes from is rarely a problem.

    It bites in two places. Confidential conversations: a customer, someone in your care, a case file, an executive board meeting, or a lawyer bound by confidentiality. And organisations with hard requirements around sovereignty or the GDPR.

    If you run conversations through a general chatbot rather than a notetaking tool, first read whether ChatGPT is GDPR compliant.

    A European stack of your own has downsides too: fewer integrations, no worldwide ecosystem of plugins, a smaller company behind it. This isn't a story of good against evil. It's a choice that depends on what gets said in the conversation.

    With a confidential conversation like that, the choice weighs heavily. That's where you want the report without your data leaving the EU. That's what Notuly delivers.

    Choosing

    Which AI suits which conversation, if your data can't go to the US?

    In two of the five rows, something other than Notuly fits.

    Which AI suits which conversation, if your data can't go to the US, as of 5 November 2026
    Situation What fits Why
    Public texts, or a brainstorm without names and figures A mainstream AI tool, an American one included There's nothing in it you'd mind leaving the building.
    An internal meeting about customers, figures or staff An AI notetaker that processes your conversation in Europe and keeps nothing of it after the report, like Notuly Your conversation is processed in Europe. Once the report has been sent, nothing stays behind at Notuly.
    A confidential conversation with a customer or about a case file, in person or online Notuly, at the table in the app or with the desktop app for Teams, Zoom or Google Meet The audio is deleted within a minute of processing, and the desktop app works without a bot in the meeting.
    An organisation that wants the processing set down in a contract, such as a local authority Notuly on Team and above, or another tool that sets down its sub-processors in a data processing agreement At Notuly the data processing agreement is available on Team and above.
    Summarising a document or PDF An EU tool for documents, not Notuly Notuly does its own recording; you can't put a document into it.

    How Notuly shows it can be done.

    The proof is a product made that way: Notuly. The AI notetaker for every conversation. Discreet in-person, hybrid and online.

    Here's what that looks like on a Tuesday. Half past two, a conversation at the table: someone starts the recording in the app, and the conversation simply carries on. For meetings on Teams, Zoom or Google Meet, the desktop app for macOS and Windows does the same, without a bot in the meeting.

    Afterwards, the report is in everyone's inbox within ten minutes: the summary, the decisions and the action points with a name and a moment ("Sam sends the proposal to Ben on Friday", "Laila schedules the follow-up for Tuesday"). The transcript comes along as a .txt file by default; you can switch it off. Notuly understands 90+ spoken languages; the report comes in Dutch or English.

    What the reader doesn't see is the route underneath. Your conversation is processed in Europe, and the whole AI step runs on our own, self-hosted models on servers in Amsterdam. Data never leaves the EU.

    The audio is deleted within a minute of processing, and your conversations are not used as training data. Your inbox is your archive. Nothing stays behind at Notuly.

    On Team and above, that comes with the data processing agreement: the document that contractually sets down the sub-processors in that chain (Article 28(2) and (4) GDPR). How we handle your data is set out in black and white on our privacy page. That's route B, from start to finish.

    For whoever has to sign this off.

    In short, for whoever has to assess this on privacy, GDPR and sovereignty:

    • Where is it processed? Your conversation is processed in Europe. The whole AI step runs on our own, self-hosted models in Amsterdam, and your data never leaves the EU.
    • What stays behind at Notuly after the report? Nothing. The report sits in your own inbox.
    • Is anything trained on it? No, never. Your conversations are not used as training data and are not for sale.
    • Do we get a data processing agreement? Yes, available on Team and above (GDPR Article 28). Working on your own with Pro? Then you can get the data processing agreement on request via team@notuly.nl.
    • What happens to the recording? The audio is deleted within a minute of processing.

    What it costs is covered in the frequently asked questions below. Still weighing the decision? Read how to check a notetaking tool upfront on privacy and the GDPR.

    Read how Notuly keeps notetaking secure →

    Sources.

    Based on public sources, consulted in June 2026 and checked again for this revision. Legislation and company policy change; check the source for the current position. This is general guidance, not legal advice.

    Frequently asked questions.

    Our IT team doesn't want conversation data ending up in the US. Which AI notetakers keep everything within Europe?+

    AI notetakers that keep everything within Europe process the conversation in Europe from start to finish, from the speech-to-text to the report. Under the bonnet, many tools use a model or cloud from OpenAI, Microsoft or Google; which ones is set out in the list of sub-processors in their data processing agreement, so ask for it. Notuly is one of them: your conversation is processed in Europe, and once your report has been sent, nothing stays behind at Notuly.

    Does the AI processing happen in Europe too, or just the storage?+

    That varies from tool to tool, and it's exactly what you should ask: an EU region for storage says nothing about where the model that processes your conversation runs. At Notuly the whole AI step runs on our own models, on servers in Amsterdam, and data never leaves the EU. Notuly doesn't store your conversations: the report sits in your own inbox. For any other tool, ask about the sub-processors in the data processing agreement.

    What is the CLOUD Act?+

    The CLOUD Act (Clarifying Lawful Overseas Use of Data Act) is a US law that obliges American providers to hand data over to the US government, even when that data sits in Europe. An American company falls under it, no matter where its servers are. At Notuly your conversation is processed in Europe, and once your report has been sent, nothing stays behind: the report sits in your own inbox.

    Is an EU cloud enough to stop your data going to the US?+

    Not automatically: an EU data centre determines where your data sits, not which law applies to the company. An American parent company still falls under the CLOUD Act, even with a server in Amsterdam. So also look at what stays with the provider after the conversation: at Notuly nothing does, because the audio is deleted within a minute of processing and the report sits in your own inbox.

    What is digital sovereignty in AI?+

    Digital sovereignty means that you, and not a foreign party, keep the final say over your data and the technology beneath it. With AI it covers the whole chain: the model, the cloud and the company with its jurisdiction. With Notuly you keep that say over the report as well: you decide who gets it. Your inbox is your archive. Nothing stays behind at Notuly.

    Do AI tools use my conversations to train their models?+

    That varies from tool to tool: with free consumer AI, training on your input is often on by default, with an opt-out, and with business services it's more often off. Always check the tool's privacy policy and data processing agreement. At Notuly the answer is: no, never. Your conversations are not used as training data and are not for sale.

    What is sovereignty-washing?+

    Sovereignty-washing is dressing up an American stack as European, with an EU data centre, an EU flag and a certificate, while the parent company and the ultimate say stay American. It doesn't settle the jurisdiction question, because that question is about the company and its partners, not about where the server is. So look at what happens to your conversation, not at the label: at Notuly the whole AI step runs on our own models, on servers in Amsterdam, and nothing stays behind after the report.

    How do I know whether an AI tool sends my data to the US?+

    Ask three questions: where is the conversation processed, what stays with the provider after the report, and is anything trained on it? Also ask for the data processing agreement, because it lists the sub-processors. At Notuly the answer is: your conversation is processed in Europe, nothing stays behind, and your conversations are not used as training data.

    Is there a Dutch AI that keeps your data in the EU?+

    Yes. Notuly is a Dutch AI notetaker whose whole AI step runs on its own, self-hosted models in Amsterdam. Data never leaves the EU. It works for in-person, hybrid and online conversations: with the app at the table, or with the desktop app for macOS and Windows for meetings on Teams, Zoom or Google Meet.

    What does an AI notetaker that stays in the EU cost?+

    With Notuly you start free with three conversations a month, no payment details needed. Above that, Pro gives you unlimited conversations for a fixed monthly price, and teams pay a per-licence price with one invoice. The current prices are on the pricing page.

    Terms

    Terms for AI that keeps your data out of the US.

    Jurisdiction
    Jurisdiction is the law a company falls under. For an AI service, it decides who can legally reach your data, regardless of where the server is.
    Sub-processor
    A sub-processor is a party a provider brings in to process your data, such as a cloud or another company's language model. A data processing agreement lists which sub-processors there are.
    Sovereign AI
    Sovereign AI is an AI service over which you, and not a foreign party, keep the final say: over where your conversation is processed, who can reach it and what's left of it afterwards.
    Self-hosted model
    A self-hosted model is an AI model the provider runs on its own servers, instead of calling it through another company. At Notuly the whole AI step runs this way, on servers in Amsterdam.

    Your conversation, processed in Europe.

    Record a conversation and see what's waiting in your inbox afterwards. The report is created on our own models in Amsterdam, the audio is deleted within a minute of processing and your data never leaves the EU.

    Three conversations freeNo payment details needed