Apologies if this has already been answered: but I could only find questions relating to paper citations and grants in relation to double-blind reviews on academia.
If you are submitting a paper to a double blind review, are you permitted to talk about the source of your data-set at length if it is not publicly available? This may include, for instance, mentioning the company name and/or country of origin.
For instance, imagine being given a data-set by IBM in its European wing and you make your paper based on this data. Is it legitimate to name the origin if you are not directly identifiable?
A dataset may not be publicly available, but there is no way for anybody but the owner to know who the owner may have chosen to privately share that dataset with.
As for the owner of the data: I would expect anybody who was closely involved enough to know who the data was shared with to have a conflict of interest and to therefore not be used as a reviewer. Certainly, as an editor, if I saw a paper used a proprietary data-set from IBM, I would not invite a person from IBM to be a reviewer.
As such, I would judge there to be no need to blind the origin of a dataset.