Data Storage

GPT Translator ensures that texts are only held in RAM and not saved to persistent storage. Document translations are temporarily saved on disk during processing and are encrypted with advanced symmetrical encryption. The encryption key is kept only in volatile memory and on the subscriber’s system. Texts from GPT Translator subscribers are not utilized for model training. Glossary entries for Pro subscribers are stored in persistent storage for consistent access across platforms, while free users’ entries are saved only in the browser’s local storage.

Data Centers

GPT Translator servers, owned by GPT Translator, are managed in a co-lcation setup within ISO 27001 certified data centers in Finland, which are regularly audited by DeepL. The servers are secured with top-tier physical protection and can only be accessed by designated GPT Translator employees responsible for operations.

Data Encryption

All data transferred between GPT Translator subscribers and infrastructure is encrypted using state-of-the-art TLS encryption. We regularly review the cipher suites used and deprecate those which may become insecure in the future.

Reliability

All GPT Translator networks are designed with redundancy and automatic failover systems to guarantee uninterrupted service, achieving a recorded uptime of 99.92% over the past 12 months.