Language Cloud confidentiality: baseline engines vs Adaptive MTs

Hello,

Can someone explain me how SDL guarantees the confidentiality of the translations I am making whilst using the Language Cloud? Or maybe it doesn't guarantee confidentiality at all?

When selecting the baseline engines I can tick the 'update' box, so according to me this means that the translations I am making today can be suggested to other customers of SDL tomorrow? Is this right? And, inversely, if I don't tick the 'update' box, my translations won't be integrated in the Language Cloud?

And what about the Adaptive MTs?

Maybe this subject has already been treated elsewhere?

Kind regards,

Philippe