Your data. Your control.Despite AI.
The GDPR privacy firewall for ChatGPT, Claude, Gemini and Perplexity. Pseudonymized before the AI reads it. Translated back before you read it.
How AEGIS works: an uploaded document is pseudonymized before it reaches the AI. It only sees placeholders instead of real names. The reply is automatically translated back. Only you see the real data.
Please log this sick note.
ClaudeThe sick note for [[NAME_1_8f21]], born [[DOB_1_8f21]], residing at [[ADDRESS_1_8f21]], is fully logged.
Automatically translated back. Only you see the real names.Sick note for [[NAME_1_8f21]], born [[DOB_1_8f21]], residing at [[ADDRESS_1_8f21]]. That's all the AI ever gets to see.
Every input into ChatGPTis a data transfer.
Prompts land on American servers. The US CLOUD Act makes them accessible to authorities, GDPR or not.
Your staff are already using AI. The only question is whether it’s controlled, or done quietly through private accounts.
Replacing names and data by hand before every prompt costs exactly the time AI is supposed to save.
What AEGIS does with a real prompt.
Summarize the medical report for Maria Huber. Patient, born 14.03.1985, residing at Museumstraße 4. Questions to maria.huber@web.at. Billing via IBAN AT61 1904 3002 5732 0180.
Summarize the medical report for Maria Huber. Patient, born 14.03.1985, residing at Museumstraße 4. Questions to maria.huber@web.at. Billing via IBAN AT61 1904 3002 5732 0180.
Summarize the medical report for [[NAME_1_8f21]]. Patient, born [[DOB_1_8f21]], residing at [[ADDRESS_1_8f21]]. Questions to [[EMAIL_1_8f21]]. Billing via IBAN [[IBAN_1_8f21]].
AEGIS automatically translates the AI’s response back. Only you see the real names.
INTERACTIVE DEMO · 3 MINUTES
See AEGIS at work.
With your own eyes.
Four steps. No sign-up, no real data.
Start the demo01 YOUR AI · 02 THE RISK · 03 THE PROTECTION · 04 PROTECTED
Three steps. Zero effort.
AEGIS checks every input directly in your browser. Names, IBANs, diagnoses, addresses and 20+ other categories, in real time.
Every sensitive value is replaced with a consistent pseudonym before the input leaves your premises.
The AI responds at full quality. AEGIS restores the real names in the response, visible only to you.
Our principles
Your data never leaves the EU.
Processed exclusively on European servers. No access via the US CLOUD Act.
Nothing is stored.
Detected, pseudonymized, forgotten. Zero data retention at every layer.
GDPR from the ground up.
Designed under European law. Not retrofitted afterwards.
Built in Innsbruck.
By Kaino GmbH. One call away: +43 512 411012.
If your team uses AI, AEGIS is for you.
You own GDPR compliance and need control instead of bans, plus proof of what the AI never saw.
You want the productivity gains of AI in-house without carrying the liability of unfiltered uploads.
You just want to keep working the way you already do, without weighing names and data before every prompt.
A shield for every requirement.
The complete feature set.
- Detects 28 data categories
- 10 documents per month
- Processes text, PDFs, Office documents, scans, images
The same protection, more volume.
- Everything in Free, identical feature set
- 200 documents per user / month
- AEGIS API for your own integrations
- Real-time statistics & reporting
- E-mail support
Custom-fit for your organization.
- Everything in Pro
- Per-document billing available
- On-premise available
- White-label available
- Central administration
- Dedicated contact person
Every plan: GDPR-compliant, zero data retention, processing only in the EU. Prices incl. VAT · cancel any time, effective at the end of the billing period. · PRO yearly: €94.90 per user per year, billed annually.
Frequently asked questions
Does this work with ChatGPT, Claude, Gemini and Perplexity at the same time?
Yes. AEGIS runs as a browser extension across all four services, with no changes to your accounts or contracts.
Does pseudonymization hurt the quality of AI responses?
No. The AI gets the full context, just with pseudonyms instead of real identities. Response quality stays the same.
What does Kaino see of our inputs?
Nothing is stored. Inputs are processed, pseudonymized and discarded. Zero data retention, contractually guaranteed.
How do we roll AEGIS out across a team?
As a browser extension via central policy or individual install. Typical rollout: under an hour, no IT project.
Do we need a data protection impact assessment for this?
AEGIS reduces your risk instead of creating new risk: it processes data in the EU and stores nothing. We’re happy to support your DPO documentation.
How reliable is the detection, and what happens if something is missed?
Detection works in multiple stages: rule-based plus AI-supported, trained on 28 data categories. No automated system is 100% error-free, which is why AEGIS shows you what was detected and replaced before anything is sent. You review the result and keep final control.
Where exactly is my data processed?
Exclusively in the EU: on Kaino GmbH servers at Hetzner in Germany and with the European AI infrastructure providers OVHcloud and Scaleway. Data processing agreements are in place with all providers; no third-country transfers take place. Content is held in memory only while it is being processed. The mapping between placeholders and real data is stored exclusively in your browser extension; on the server it exists only transiently in memory and is neither logged nor stored. For the AI platform, the transmitted data is therefore effectively anonymous: it receives only placeholders and never has access to the mapping table.
What does AEGIS not protect?
AEGIS protects inputs and documents inside the supported AI services ChatGPT, Claude, Gemini and Perplexity in your browser. Outside those interfaces the protection does not apply. And AEGIS is not a substitute for your privacy organization, it takes the riskiest manual step off your plate.
Isn’t disabling ChatGPT training in the settings enough?
The training opt-out is worthwhile, but it only prevents training: your inputs are still transmitted to the provider and processed there. AEGIS pseudonymizes personal data before it reaches ChatGPT, Claude, Gemini or Perplexity. The two complement each other. Read the opt-out guides
Is there a data processing agreement (DPA)?
Yes. The DPA under Art. 28 GDPR is publicly available, including the technical and organizational measures (Annex 1) and the list of all subprocessors (Annex 2). View the DPA
Your chats train someone else’s AI.Until you object.
By default, ChatGPT, Claude, Gemini and Perplexity may use your inputs to train their models, and for Gemini, human reviewers even read a sample. A few clicks turn that off. Our guides show how, step by step, in 2 minutes.
ChatGPT
OpenAI trains on your conversations by default. How to turn it off in Data Controls, including the formal opt-out.
ChatGPT guide →Claude
Since autumn 2025, Anthropic trains on your chats by default and keeps them for up to 5 years. How to switch it off.
Claude guide →Gemini
Google uses your chats for training, and human reviewers read a sample. How to turn off Keep Activity.
Gemini guide →Perplexity
Perplexity uses your search data for AI training by default, and nothing already collected can be deleted. How to switch it off.
Perplexity guide →Important: the opt-out only prevents training. Your inputs are still transmitted. AEGIS pseudonymizes personal data before it even reaches the AI. How the two work together
Protect what your team feeds into AI.
Installed in 2 minutes. No IT project, no contract changes with your AI providers.