GDPR, AI Act & data security
Frida reads your customer emails to build your sales overview. That is why data protection, human control and AI transparency are not features for us — they are the foundation.
Processing primarily in the EU — hosting in Germany/Finland
AI Act transparency — users know when AI is making suggestions
Disclosed sub-processors and transfer basis — no hidden links
Your emails are never used to train AI models
Frida is built EU-first. Hosting and storage run on EU servers, and email AI processing happens primarily on EU-based infrastructure.
Every sub-processor we use is listed openly in our privacy policy. Under peak load, AI processing can overflow to a US-based sub-processor — always under the EU-US Data Privacy Framework and standard contractual clauses, and your data is never used to train models.
The AI Act is partly about transparency, responsible use and human control. Frida is built around one principle: users should know when AI is making suggestions, and nothing leaves the inbox without approval.
GDPR is not only about where data lives. It is about control. Frida makes it clear what is processed, why it is processed and how you can have it deleted.
Frida is developed by a Danish company under EU law. We built the product for businesses that take their customers’ trust seriously — because our own product depends on exactly that trust.
Comparison
Many AI tools are excellent — but customer emails are personal data, and the difference lies in the paperwork and the defaults. Always check the terms yourself.
Yes — listed in the privacy policy.
Often unclear on free tiers.
Primarily in the EU, with disclosed overflow under EU-US DPF.
Often undisclosed, typically US.
Never.
Often the default unless you opt out.
AI suggestions are shown as AI suggestions, and the user stays in control.
Varies — especially in generic tools.
You — every message requires your approval.
Varies.
Full deletion on request.
Varies — check the terms.
FAQ
Yes — when processing is necessary to provide the service, data processing is limited, sub-processors are disclosed, transfers are safeguarded and the user stays in control. Frida is built around those principles.
No. Your data is used to deliver the service to you — drafting replies and building your overview. It is never used to train models.
Primarily, yes: hosting and email AI run on EU-based infrastructure. Under peak load, processing can overflow to a US-based sub-processor under the EU-US Data Privacy Framework — all sub-processors are disclosed in the privacy policy.
The list is in the privacy policy with region and transfer basis. A separate data processing agreement is not part of the standard terms yet.
Frida is built with AI Act transparency and human control: users know when Frida is making AI suggestions, and drafts or replies are not sent without approval. We continuously document AI purposes, providers and risks so the product can follow the relevant requirements as they come into force.
Yes. You can request full deletion at any time, and deletion follows our published retention policy.
Frida is an AI-powered CRM for Outlook that automatically creates and updates deals from email conversations. It also drafts replies, tracks follow-ups, and keeps customer context connected to the inbox.
Yes. Frida is built around Outlook and Microsoft 365, so teams can keep working from the inbox they already use.
No. Frida prepares drafts and suggestions, but the user stays in control and approves messages before they are sent.
The category Frida wants to own
Frida is an AI-powered CRM for Outlook that automatically creates and updates deals from email conversations.
