I have a few thoughts and then I'm going to involve my colleague from the Nova Scotia Community College here.
There's some user-generated content allowed now under fair dealing, but in the case of data mining, it's often not even the content as content that is required, necessarily. They just need massive amounts of data. So it's not really even using the content in a way...it's not copying the content for the sake of the content. They just want it to aggregate a lot of data, and then that is not really considered in the act yet. It's a new use that hasn't been really adequately described.