Security & privacy
Security,
Security,
by design.
Your data in the EU, human control, total traceability. Sania is built for confidence — not just performance.
✓Control
SAN
SAN
The Swiss Agent Network control framework
Our internal framework covers permissions, traceability and human control. It does not constitute a certification issued by a third party. You know what Sania does, what she accesses and what actions await your validation.
Security, by design
Six non-negotiable principles, applied to each agent. This framework – not the model itself – is what makes secure AI in Switzerland, able to work with private company data.
🇪🇺
Data in the European Union
Your data is hosted in the European Union: the application in France, the database and file storage in Ireland. The processing locations, subcontractors and applicable supporting documents are documented in the contractual framework.
🚫
No reuse for training
We do not reuse your content to train models, and we select providers that offer guarantees against reusing your data for training purposes.
✅
Approval inbox
Sensitive actions go through your validation before execution. Humans keep control of what matters.
🔍
Review of unmanaged AI tools
We review with you the unmanaged AI tools used in your organisation and frame their use. This is an internal review, not an audit certified by a third party.
📜
Total traceability
Each Sania action is logged and viewable in your dashboard. No black box.
🔐
Access & permissions
Sania only accesses the services and data that you explicitly authorize, service by service.
View ourprivacy policy. DPA, accommodation details and applicable supporting documents are available on request at contact [chez] swissagentnetwork.com.
Security
AI that
you can trust.
EU hosting. Human control. Zero black box.