Brainstorming Intelligence Reform

Senior U.S. intelligence officials meeting in Denver on Monday revealed new measures to try to fix the nation’s ailing intelligence system, which insiders say leaves the government overloaded with data but unable to answer key questions.

One leader compared the 16-agency U.S. intelligence community to 8-year-old soccer players bunching around the ball with much of the field uncovered. That could mean potential deadly misreads on terrorism, Iraq, Iran, China and North Korea.

U.S. intelligence collectors “have become vacuum cleaners on steroids,” said Thomas Fingar, deputy director of national intelligence, at the opening of the four-day Information Sharing Conference.

Yet intelligence analysts - 50 percent of them with less than five years of experience - are mired in uncertainty over who should have access to what, Fingar said.

Find me the honest analyst that maintains that sentiment and I’ll show you a political animal angling for SES. The only people fretting about what to share are those with budget dollar signs in the eyes and dreams of building their own ricebowl or taking over someone elses.

The result is vast amounts of unanalyzed information that “is just data. We are awash in data,” he said, calling for a system-wide revolution.

“Otherwise, we become a very expensive irrelevance. … If we don’t believe it is doable, we all ought to resign,” he told several hundred conference participants at the Hyatt Regency Denver at Colorado Convention Center.

Among the reform measures [being planned]:

• Launching an experimental “Wikipedia”-style approach to a major intelligence challenge - understanding the situation in Nigeria - in which all analysts across the 16 intelligence agencies are invited to contribute what they know, as the public does at the online free-encyclopedia site Wikipedia. The idea is to see how much can be learned across agency lines, not just among analysts with Nigeria as part of their portfolio.

• Creating a new standard format for intelligence reports that all agencies would use so that every analyst in the intelligence community can have access to all the information. The new format would give place, time and a rating on source credibility, stripping away pages of routing information and source details that can impede analysis.

• Mobilizing a “geek squad” to persuade veteran analysts to embrace new technology. For example, rather than jot notes in margins of intelligence report printouts and then file them away, veterans adept at asking the right questions would jot notes electronically and then file the annotated documents in a common- access electronic folder.

• Building a database listing all analysts and their expertise so leaders can know who does what. Then they can form teams across agency boundaries. An estimated 100,000 people work in the U.S. intelligence community, the bulk of them in the Defense Department.

• Assigning analysts at two agencies to answer the same question, such as an assessment of Iraq’s new government, with one team using widely available open-source information and the other using classified material. […]

God, where to start . . .

The intel cycle is of course the root of most of the evils here. People who harp on the gatekeeping that goes on in the MSM would stroke out if they were aware of the hoops one has to jump through to ask a simple question and get back an answer. Things are so slow you might not know you have a chance to ask a question until after the opportunity is gone. Culling collection and production management staff and building a “marketplace” for tasking and response that puts collectors and analysts closer together is a start. Build in Ebay-style rankings and you’ve got both your refined collection-production cycle and a mechanism to evaluate performance at the end of the year.

Moving to a blogging environment cuts out the frivolous system of hardcopy, glossy color production and the support staff that keeps it all alive. Not that we don’t need editors (God, did I need an editor) but when you can put your assessments out in real- or near-real time why be forced to slip customers a bootleg copy while they wait a month or more  for the “official” version? NIC meetings? Analytic exchanges? Ridiculous when everyone is boucing ideas and vetting thoughts 24/7. Political manipulation? Unlikely when all your work is “public.” Build in a Digg-style rating system and you’ve got not only a way to push the best analysis to the front, but (once again) a way to evaluate performance at the end of the year.

The Wiki-idea is another good start and should have been implemented much earlier. That they’re merely dabling with Nigeria (not a ‘nothing’ problem but not exactly front-page material) is telling. There is no reason why they should implement it community-wide now for all subjects, save for fear that those pesky aforementioned gatekeepers would be cut out of the loop.

The new report format idea is a make-work snow-job. Crafting a script to strip away all admin overhead in reports is a trivial exercise. In fact if I’m not mistaken I think I did the very thing in roughly ‘94 with my scary TA-270 skillz.

Frankly I don’t put a lot of stock in the “geek squad” idea. Look, I knew guys who still worked with pencil, paper, magnifying glass (yes, he was old) and guhor stick when the IBM XT was the standard computer workstation in the IC. In a sense things haven’t gotten better for that crew; it has gotten worse. By all means tap their expertise, but don’t burn a lot of cash thinking you’re going to turn them into power users. Retire them, put them in an IC Reserve, and let them do what they really want to do: dig deep, think big thoughts, write down big ideas.

The expertise database is another BS suggestion. Find me the agency that doesn’t have one already, and let’s not forget the data call for just such info that came down in, what, ‘03? There is also the NIPF procress, which puts all the relevent SMEs together several times a year. It isn’t a matter of not knowing who to call, but breaking down the barriers that preclude their effective cooperation.

The “Burundi Experiment” where OSINT competes with other -INTs is nice theater but that battle was fought and won. A more significant step forward would be to require that every assessment start with a base of OSINT and a sprinkling of other -INTs when they are necessary to complete the picture. Not only would this show how cheaply we could run the IC, it helps facilitate all that State and Local-level sharing everyone keeps saying they’re going to do, and makes declassification/downgrading for wider public consumption that much easier when the next controversy comes down the pike (like you know it will).