At National Defence, the security of information and intelligence is really important. In my opinion, the problem starts when we have to establish the security classification of a document for the first time.
This is why training is so important. We have a bias in the department of trying to classify everything as secret and protected right from the get-go. Once something is classified that way, any ATIP request that comes in requires more care and attention to process.
If you train people properly on what truly is secret and what's protected and if you classify things the right way the first time, that leads to greater efficiencies. However, without proper training, I think the bias in a place like the Department of National Defence—because security is paramount—is going to be to overclassify documents from the start. I think we could do more when a document is first created to say that it shouldn't be classified. That would alleviate part of the problem.
The other piece is that now, under open government, we are going through and trying to declassify a bunch of documents and release our datasets.
The work is ongoing, but once the classification level is established for those documents, a lot of paperwork has to be done before information can be published.