Back to Home

Frequently Asked Questions

1. Why EU GPT Exists

Why is Sovereign AI a strategic necessity?

Because AI only works if your internal data is processed inside the provider’s infrastructure. If that infrastructure falls under US or other non-European law, your data is no longer fully under your control. EU GPT is built based on a European provider, ensuring that intellectual property, sensitive information, and business intelligence remain protected within the European legal sphere.

What is the strategic risk of using U.S.-linked AI platforms?

U.S.-linked AI providers are subject to American laws that override European legal safeguards. This means your data can be accessed outside EU judicial oversight, regardless of contracts or compliance claims. EU GPT removes this structural risk by operating entirely under European legal authority.

Why does the U.S. CLOUD Act matter?

The U.S. CLOUD Act allows American authorities to demand data from U.S.-controlled companies regardless of where the data is stored. Physical location offers no protection, legal jurisdiction does. EU GPT operates entirely outside CLOUD Act reach.

What is FISA Section 702 and why does it matter?

FISA Section 702 allows U.S. authorities to force American companies to provide access to data of non-U.S. citizens, without a warrant and outside European legal procedures. This access is secret, mandatory, and cannot be challenged under EU law. EU GPT is not exposed to FISA 702.

How has FISA Section 702 expanded recently?

The 2024 expansion significantly broadened its scope, allowing U.S. authorities to compel almost any entity with access to data or systems to assist in surveillance activities. Even indirect technical access can trigger these obligations.

What can U.S. cloud and AI providers legally do with your data?

U.S.-linked cloud and AI providers may legally be required, or allowed, to reuse, further process, or make data accessible for purposes beyond the original intent, including system improvement, monitoring, or model retraining. These activities can occur outside European legal oversight, even when contracts claim otherwise.

What is Data Colonization in AI?

When European organizations rely on U.S.-controlled AI, their knowledge and intellectual property effectively leave the European legal sphere. This strengthens foreign AI ecosystems at Europe’s expense. EU GPT prevents this by keeping European knowledge under European control.

What is Sovereign-Washing?

Many platforms claim to be European while relying on U.S. technology, ownership, or control behind the scenes. Despite European branding, U.S. law still applies. EU GPT delivers real sovereignty, not sovereign-washing.

2. Privacy, Data Control & Safe AI Usage

Are my documents or conversations shared with big tech?

No. Many AI platforms forward data to U.S.-controlled models. EU GPT processes data exclusively within its own sovereign environment. Documents are used only for the active session and are never shared or used for external model training.

Who owns the data?

You do. You remain the 100% owner at all times. While U.S. providers can be legally forced to provide access, EU GPT is built on a Sovereign-by-Design principle, with no foreign legal backdoors.

What happens to my data when I use EU GPT?

Your data is processed only to deliver your requested output. It is not reused, not retained, not shared, and not used for training external models.

How does EU GPT ensure privacy-compliant AI usage?

EU GPT ensures that data:

  • Can be processed only for the intended purpose
  • Is not reused, shared, or repurposed
  • Remains under European legal control
How does EU GPT expand the scope of AI applications safely?

Because EU GPT operates fully within European privacy and security boundaries, organizations can safely use AI with privacy-sensitive and confidential data.

Why is this especially important for privacy-critical organizations?

For organizations in sectors such as public, healthcare, legal, finance, energy and pharma sectors, privacy risks often prevent meaningful AI adoption. EU GPT removes this obstacle.

Does this also apply to individual users?

Yes. The same principles apply to individual users that provide sensitive information or want to use EU GPT following its European origin.

What is the practical impact of this approach?

It enables AI to be used where it was previously considered too risky. Privacy becomes an enabler instead of a limitation.

3. Security, Compliance & Data Processing

Why does EU GPT use Scaleway as Cloud Provider?

EU GPT uses Scaleway because it is a fully European, sovereign cloud provider designed for high-security and government-grade use cases.

How secure is Scaleway?

Scaleway operates at SecNumCloud-level security, the highest French/European security classification, comparable to the highest Dutch BIO, BBN and BVC levels.

What are BIO, BBN and BVC and why do they matter?

  • BIO (‘Baseline Informatiebeveiliging Overheid’) determines whether sensitive data may be processed at all
  • BBN (‘Basis Beveiligings Niveau’) determines how sensitive that data may be
  • BVC (‘Beschikbaarheid, Vertrouwelijkheid en Integriteit’) defines availability, confidentiality, and integrity requirements

EU GPT is designed to operate at the highest levels across these frameworks.

Why do U.S.-linked platforms has the lowest BBN and BVC levels?

Because foreign laws like Cloud Act and Fisa create mandatory legal backdoors that cannot be mitigated through technical or contractual measures.

What makes SecNumCloud-level security different?

It includes hundreds of strict security controls, continuous national supervision, protection against advanced persistent threats, and immunity from extraterritorial laws.

Does EU GPT add security beyond the cloud provider Scaleway?

Yes. EU GPT applies additional layers such as strict access controls, audit trails, encryption, least-privilege access, and secure development practices.

What happens to my data technically?

Your data is not reused, not shared, and not used for training of the AI models. Your data is processed in real time. It will be maintained only within the chat function. If deleting the chat your information will be automatically deleted. In case of uploading documents you have full control of these documents within a central overview and can be deleted at your discretion.

4. AI Models, Technology & Transparency

Which AI models does EU GPT use?

EU GPT uses state-of-the-art open-source Large Language Models under permissive licenses. The team selects the best performing general purpose models based on their extensive AI experience. This involves models with 200 to 400 billion parameters for the optimal combination of speed and accuracy.

Which model should I use?

Reasoner is optimized for knowledge and logic. Choose this model for complex questions, factual accuracy, and deeper analysis. It possesses extensive internal knowledge and strong reasoning capabilities.

Searcher is optimized for retrieval and web access. Choose this model when you need to find specific information within your documents or the live web. It acts as a navigator to locate data but relies heavily on external sources for its answers.

Why open-source AI models?

They are transparent, auditable, mature, and do not store data or learn from user interactions.

How does EU GPT achieve high performance?

Through the selection of modern efficiency techniques that activate only relevant parts of the model per request, resulting in fast responses, high accuracy, and scalability.

5. Customer-Centric Value Delivery

What does EU GPT offer?

  • Organization silo set up for highest security and control
  • Two different high performing Large Language models
  • Custom GPT/workspaces
  • Source reference and document page numbering
  • Control on uploaded information by overviews of chats and files.
  • Web search
  • Strong mathematics capabilities
  • Separate spelling LLM
  • Prompt suggestion
  • Front-end
Who is EU GPT for?

EU GPT is designed for professionals and organizations operating in privacy-critical environments such as public, healthcare, legal, finance, energy and pharma.

What is EU GPT mainly used for?

Based on research of more than 200 use cases across various organizations, EU GPT focuses on professional, information-centric work such as content creation, rewriting, internal knowledge management, audits, compliance reviews, and document analysis. Implementation by our experienced team consistently delivers over 90% of the directly obtainable AI value for organizations.

How does EU GPT ensure customer value instead of just providing software?

EU GPT operates according to a proven Way of Working (WoW) that has already delivered substantial, measurable value within the large organizations. In short, our WoW is explicitly customer-centric and focuses on:

  • Selecting the right high-impact use cases
  • Fast, tangible value realization
  • Controlled and compliant scaling
  • Integration into existing professional workflows
EU GPT does not leave customers alone with a tool. We are available to fully service organizations, guide adoption, embed AI operationally, and ensure value is realized in practice.

6. Accuracy, Risk Management & Traceability

How does EU GPT reduce linguistic errors?

Through additional spelling, grammar, and language-quality controls tailored for professional use.

How does EU GPT support the reduction of hallucination?

By providing explicit source references and page numbering. Often LLMs perform limited on calculations. EU GPT targets to prevent this type of hallucination by including a mathematics engine. The human in the loop check on output is always recommended.

7. Organization, Ecosystem & Scaling

Who is behind EU GPT?

A fully European team with decades of combined experience in delivering production-grade AI within large organizations.

What is EU GPT’s vision on AI in the workplace?

AI should increase effectiveness, creativity, and job satisfaction in order to put their customers first.

How does EU GPT scale across Europe?

Through partners and ecosystem collaboration, enabling vertical integrations and sector-specific solutions while maintaining sovereignty, privacy, and compliance.

8. Access, Adoption & Daily Use

How do organizations get access to EU GPT?

Via sales, the EU GPT link will be provided for your domain. Registration takes place at the bottom of the page.

How does EU GPT support adoption?

Through workshops, focused pilots, and scalable roll-outs once value is proven.

What is the difference between a Chat and a Workspace?

Chats are for quick tasks and summaries of limited text. Workspaces support large knowledge bases, ongoing projects, customization, and collaboration.

9. Why Choose EU GPT

Why choose EU GPT?

Because EU GPT combines:

  • Sovereign European AI
  • Privacy-safe processing
  • Proven, customer-centric value delivery
  • Full servicing via a proven WoW
  • Scalable deployment through partners

In one sovereign & compliant platform.