A short chain of answers
Questions about where data sits and who can reach it can be answered without asking anyone else. That shows in procurement and in audits.
The whole system is developed in Sweden by us, and runs on our own servers in our own data centre. That includes the AI. No part of the chain sits with someone else.
A security system is only as independent as its weakest supplier. If the analysis sits with a cloud provider, the model with an AI provider and the operations with a third party, then those three decide what happens to the information — not whoever sold you the service.
We made the opposite choice. Code, operations, hardware and AI are ours. That keeps the chain short enough to describe on one page, and it makes the answer to "who can reach this?" a short one.
Questions about where data sits and who can reach it can be answered without asking anyone else. That shows in procurement and in audits.
The threat data and the blocking events we receive stay in our own environment. There is no third-country transfer to risk assess.
If something needs adjusting — a rule, an interface, a format — we do it. We are not waiting on a supplier's roadmap.
We use AI to find patterns in the threat data coming in and to summarise the situation for those working in the system. It is reasonable technology for that task.
But the model runs on our own hardware in our own facility. No threat data, no logs and no customer details are sent to an external AI service to be analysed. That is a technical constraint in how the system is built, not a policy that could quietly change.
If you need documentation for a supplier assessment or a procurement — a description of the operating environment, where data is processed, which sub-processors are involved — we will send it on request. We prefer answering written questions concretely to pointing at a page of generalities.
Describe what you need to know and we will answer concretely.