Skip to content
KameleonAI

Why offline?

Your data belongs at home. Not on a trip.

A plain-language explanation: why sensitive data should never leave the building, why that is finally possible, and why we think this is only the beginning.

The simple idea

You wouldn’t keep your filing cabinet at the neighbours’, would you?

Picture it: all your case files, contracts and patient records in one filing cabinet. You wouldn’t put that cabinet at an unknown company on the other side of the ocean. You keep it at the office, under lock and key.

With an online AI service, your questions and the documents you attach are processed on a provider’s infrastructure. Where the data sits, how long it is kept and who can access it varies per service and per contract.

Kameleon turns that around. The AI comes to your data, instead of the other way round.

You wouldn’t keep your filing cabinet at the neighbours’, would you?
You wouldn’t keep your filing cabinet at the neighbours’, would you?Online AI serviceKameleon
Where is your data?In somebody else’s data centreIn your own office
Who runs the environment?The provider, under its contract and applicable lawYour organisation
Works without the internet?Usually notYes, over the local network
Who sets the price?The provider, every year againOne-off, plus a fixed maintenance fee
Is your data used for training?Depends on the provider and the agreementNot by any external AI service

Who’s looking over your shoulder?

Three things you should know

01

The American CLOUD Act

The CLOUD Act of 2018 gives US authorities, under certain conditions, a route to request data held by American providers, even when that data is processed outside the US. Which procedure and safeguards apply depends on the request and the situation.

02

The Chinese intelligence law

Providers outside the EU may be subject to different rules on government access and data processing. So always check where data is processed, what has been agreed and which law applies.

03

The cloud is just someone else’s computer

Even a well-run cloud service remains an external dependency for availability, terms and data processing. After Schrems II (2020), the EU adopted a new EU-US adequacy decision in 2023 under the Data Privacy Framework. That provides a legal basis for transfers to certified organisations; each processing activity still needs a careful assessment.

This is a simplified explanation, not legal advice. Want to know exactly how it applies to your situation? Ask us and we’ll put you in touch with a specialist.

The bigger picture

How we got here (and where we’re going)

Digital sovereignty sounds like a word from a policy paper. It actually means something very simple: being in charge of your own data and technology.

  1. 2006

    The cloud begins

    The first big cloud services launch. Renting servers instead of buying them: easy and cheap.

  2. 2010 – 2020

    Everything moves to the cloud

    Email, documents, accounting. A handful of American companies manage more and more of Europe’s data.

  3. 2018

    The CLOUD Act

    The US passes a law regulating access to data held by American companies, wherever in the world it is stored.

  4. 2020

    Schrems II

    The European Court of Justice invalidates the Privacy Shield. Organisations have to reassess their international data transfers.

  5. 2022

    AI goes mainstream

    Chatbots become mainstream overnight. Many popular services run in the cloud of large tech companies.

  6. 2023

    EU-US Data Privacy Framework

    The EU adopts a new adequacy decision for transfers to certified American organisations: Implementing Decision (EU) 2023/1795.

    Read the official decision on EUR-Lex
  7. 2023

    Open models take off

    Companies such as France’s Mistral and Meta release open models that anyone can run themselves. The beginning of an alternative.

  8. 2024

    The European AI Act

    The European AI Act enters into force. Its rules cover transparency and accountability, among other things.

  9. Now

    The tide turns

    Open models can handle more and more everyday document work locally. That gives organisations a real choice alongside cloud services. Kameleon helps them put that choice into practice.

Our philosophy

Kameleon isn’t a gadget. It’s a way of working.

Five things we believe in. They shape what we build, and what we deliberately don’t.

  1. 1

    What’s yours stays yours

    Your data, your AI, your cabinet. We build in no back doors and no hidden subscriptions.

  2. 2

    Open over closed

    We use open models and open-source software. That makes control and portability possible and limits dependence on a single closed vendor.

  3. 3

    Local over leaky

    Local processing shrinks the online attack surface and keeps questions and documents inside your own network. For us, offline is a deliberate design choice.

  4. 4

    Explaining over impressing

    If we can’t explain it simply, we don’t understand it well enough ourselves. So: no jargon.

  5. 5

    Repairing over replacing

    A Kameleon lasts for years. Parts can be replaced, models can be updated. No throwaway hardware.

Ready to bring your AI home?

We’re happy to think it through with you. Even if you’re only just getting started.