Your AI traffic, in your environment
Unseen runs in your own data centre or cloud account. Prompts go only to the AI providers and regions you approve, and the record of every call stays with you.
Every AI call passes through your environment
People, apps and agents reach AI through Unseen, running on infrastructure you control. From there, traffic goes only where you have said it may go, and three things hold for every call.
- People
- Apps
- Agents
- VMware
- Nutanix
- AWS
- Azure
- Google Cloud
- Kubernetes
- OpenAI
- Anthropic
- Google Gemini
- Mistral
- Azure AI Foundry
- AWS Bedrock
- Your own models
In the regions you choose. Models you host yourself keep everything inside.
Sensitive data removed
Personal data, credentials and confidential content are taken out before a prompt or file leaves.
No training
Models are reached through API access that excludes model training.
The record stays with you
Who sent what, where it went and what was decided, kept in your environment.
In your data centre or your cloud account
Run Unseen on VMware or Nutanix in your own data centre, or in your own AWS, Azure or Google Cloud account. It is not a service you have to send your traffic to.
Your prompts, files and answers stay in your environment until they go to the provider you chose. Unseen's cloud never receives them.
Only the providers and regions you approve
You decide which AI providers Unseen may use and in which of their regions, for example Azure OpenAI or Mistral in the EU.
For the most sensitive work, point it at models you host yourself, and the data never leaves.
Four questions to ask of any AI service
Sovereignty is not only where the servers are. Under the US CLOUD Act, a US provider must disclose data in its possession, custody or control "regardless of whether such communication, record, or other information is located within or outside of the United States".
Where is it stored?
The country of the data centre, and whether you can pick the region. Our directory shows that most AI services do not say.
Whose law applies?
Where the company is based, not only where the servers are. A US provider can be compelled to disclose data it holds abroad.
How long is it kept?
Fixed period, until you delete it, or "as long as necessary". The last one means there is no limit you can check.
Does it train a model?
Consumer tiers often do by default. Business and API terms usually exclude it, in writing.
What AI services say about it
Questions
What is AI data sovereignty?
It is knowing, and deciding, where the data in your AI use goes: which company receives a prompt, in which country it is stored, which law can compel access to it, how long it is kept and whether it trains a model. For an organization it means those answers are set by you, not by whichever AI tool an employee picked.
Where can Unseen run?
In your own data centre on VMware or Nutanix, or in your own AWS, Azure or Google Cloud account, including on Kubernetes. It runs in an environment you control, so prompts and files pass through your infrastructure, not ours.
Does Unseen see our prompts?
No. Prompts, files and answers stay in your environment until they go to the AI provider you chose. Unseen’s cloud never receives them.
Can we keep everything inside the EU?
Yes, if the providers you approve offer EU processing. You choose which providers and which of their regions Unseen may use, for example Azure OpenAI or Mistral in an EU region, or models you host yourself so nothing leaves at all.
Does storing data in the EU protect it from US law?
Not on its own. Under the US CLOUD Act, a US provider must hand over data in its possession, custody or control regardless of whether it is stored inside or outside the United States. Where the provider is based matters as much as where the data sits.
Do the AI providers train on our data?
Unseen reaches models through API access that excludes model training. Consumer accounts are different: on a personal ChatGPT account, training is on unless the user turns it off.