HUMINT acquisition has been a problem for the CIA since the '90's which is a major problem when dealing with 7th century minded savages isthe Middle East. They got too comfortable with news journalists doing their leg work for them and publishing them inthe public domain which was a whole lot easier to do. Now there is a whole new generation that posts all over the Internet and some of that info can be helpful but the problem is you have to identify their online hangouts and translate their conversation that's written in foreign languages such as Arabic or Urdu.
The CIA needs to begin retraining their case officers more in field craft and deploy good candidates who have the appearance of the people living in the environment that they will be injected in IMHO. HUMINT is more timely, typically more credible, and more effective in removing serious threats in the Middle East. You can't identify them with open sources...
I can’t say I take exception with any of your statements except the last one, “You can’t identify them with open sources...”
I worked with retired USAR COL and former US Congressman Rob Simmons when he commanded an Army Reserve Strategic MI detachment back in the early to mid-90s; his unit provided general support to my active-duty unit on a number of our ops. Together, Rob and I developed some early methods for identifying and integrating information from the public domain - not necessarily the Internet at the time - into our operational intel, as well as that from national sources.
Our methods were based on techniques we applied to tactical and operational-level collection and a tachnique we used to call “time-event charting”, which we used to reconstruct activities leading up to insurgent and terrorist ops. As information became more ubiquitous and more international with the expansion of the ‘net, we adapted and refined our methods.
We used financial information (not bank data, but identifiable financial and business relationships), along with a lot of other information that existed in the public domain, some of it only available via subscription services (obviously, we anonymized our searches and subscriptions). A lot of those methods are now integrated into what’s come to be known as “Activities Based Intelligence” and “Object Based Production”.
Using these techniques, we identified a LOT of individuuals and cells of various types, i.e., ops, logistics, intel, financial support, etc., in the the United States and many other countries. For example, our work led to the opening of over 100 FBI CT-related investigations in the US alone.
The real challenge now is that the IC just doesn’t understand what OSINT really is. And that opinion is not limited to the CIA; it also applies to DIA, NSA, NGA and others. In my opinion, within the IC, only the FBI and DEA, for the reasons noted above, have a full appreciation for the potential value of information from the public domain, which goes way beyond the contemporary definition and understanding of “OSINT”.