Trados security FAQ
Here's the answer to this frequently asked question about security.
FAQ
Do you train machine translation models with my data?
In the Trados platform, no AI model training takes place. When content is processed by a Language Weaver Generic language pair, or by an LLM, it is only available for the duration of the query and response. Your data remains yours - it's never shared publicly or reused in any way.
In Language Weaver, customer data can be used, on request and with consent, to train private adaptive, auto-adaptive, or generative language pairs. This data is only used to train your private model and is never used for any other purpose. Once complete, all training data is immediately purged and cannot be reconstructed from the model.
Related resources
