Translation Memories and client confidentiality
Recently I found a bit of a “strange” translation in a TM while I was doing some database maintenance. My eye only fell on it because the name of a cartoon character was mentioned - which seemed pretty surprising to me as I was working on a TM of a client in the financial sector. The translation was actually a good laugh (a script of a cartoon movie). Besides having this strange feeling that I just saw something that wasn’t meant to be seen by my eyes, I actually didn’t think anything more about it. At the end of the day it turned out to be just a few fragments of a job which the translator must have done (by accident I assume) with the TM we used for this client.
But then a few days ago I found a clip on the common sense advisory web page http://www.commonsenseadvisory.com/news/global_watchtower_one.php?wat_id=254. The article mentions that SDL has asked language service providers (LSPs) to sign NDAs in which teh client is asked to acknowledge that SDL has access to their TMs and thus to potentially confidential material of the LSP’s clients.
I have to say that this is an interesting point. Besides many legal issues about the LSP’s NDAs with their clients, I wonder if this issue can be seen as an argument for document leverage rather than TM leverage. The difference between the two in this respect is that when you gain leverage from the client’s translated legacy documents, you stay within the institution you translate for. Compare that to trying to leveage from your “big” TM for all clients. You can, of course, filter by client names, but can you really make sure that no confidential information leaks out of the TM?
Even though the argument doesn’t sound very convincing: this problem definitely looks different with an XLIFF based translation approach, in which only relevant information is relayed to the translator.
In any case it is an interesting viewpoint on NDAs and TMs