Based on my experience as a Canuck academic working on US security issues, John’s explanation is spot on. My work on the US military (and other militaries) is by necessity based in part on interviews with current and past participants, so arguably in some senses my research is at the very edge of ‘open source’ (or possibly a bit over that edge). Indeed I have on occasion been asked up front whether I have clearance at the start of an interview.

The crux in John’s explanation is whether the researcher has clearance to work on classified material; if s/he does have clearance, then even their open-sourced work will be reviewed for the reason John notes, (this seems to fit Marc’s Canadian example as well), but if the researcher does not have any clearances then their work is not subject to review or classification. This has always been my understanding stemming from conversations I have had about the possibility of my shifting from ‘public’ academia to ‘military’ academia, or even to consultant work, for should I have agreed I would have had to obtain a certain level of clearance and hence my work, as it was consistently explained to me in all cases, would then be subject to review (if not, as John says, classified derivatively if my work was on classified issues/materials).

I have never gone that particular route for a variety of different reasons, but one reason is that if I ever was given clearance then my work would in the least be subject to review (but this has never been the main reason – cutting off my pony tail and wearing a suit daily always are much higher on my list of reasons ). Possibly the one gray area exception that I have run across is some work I did, based completely on my open source research, for the US gov't via a consultant; in this case I could/can disseminate what I had written but was asked not to distribute the final document in which that work appeared (in whole) without prior permission as the document was for 'internal use only' (this exception seems to sort of fit Marc's example re red teaming and self censorship).

So, all in all, concern by an academic (who works and researches solely in the public realm) about having their open source research censored, unless they have clearance and/or are working on classified issues, seems to me to be more than bit misguided.