Rep. Doug Lamborn's 15 minutes of fame begins and ends with human error. I'm betting that's the case, anyway. In a Congressional hearing, he read a paragraph from a Defense Intelligence Agency assessment of North Korea's nuclear capability:
"DIA assesses with moderate confidence the North currently has nuclear weapons capable of delivery by ballistic missiles. However, the reliability will be low."
The Chairman of the Joint Chiefs of Staff, Gen. Martin Dempsey, seemed caught off guard. Such statements don't usually make it into a public hearing. They're usually classified. Now, how could a statement like that be "reasonably expected to cause grave or serious harm to national security," which is the baseline standard for classification? The way the intelligence community would see it is this: by giving the North Koreans our assessment of what their capability is, we give them a tactical advantage, particularly if "we" are wrong; we give them a psychological advantage because they can now adjust their deception efforts to match or exceed such projections; and potentially, we reveal how the U.S. obtains its intelligence, particularly if the North Korean ballistic missile program is very closely held and well protected.
A few things to keep in mind. I don't doubt that the paragraph in question was marked with a big (U) in the report Lamborn referenced. But that does not mean that Lamborn (or anyone else) is allowed to make it public. If it's part of a classified document, the entire document is considered to be classified, even though many portions are not, and will be labeled as such, largely because of both habit and information-sharing priorities.
The existence of the document might be classified, too; every document that's classified includes information that is unclassified. In a different setting, if the intelligence community believed it had a responsibility to inform the American people (and not just policy makers) on the matters it analyzes, then it would routinely make excised versions of its estimates available. Even though I oppose overclassification as much as the next bloke, I understand why intelligence estimates are not written for public release, and are not released, except in extraordinary, usually political, circumstances.
Their intent is to provide the people who make decisions with information that allows them to make better decisions. Secrecy here is properly used as the interface that permits the strategic surprise. Those of us who have studied the secrecy apparatus and believe it to be in need of reform need to agree to a basic set of stuff that will be automatically classified because of the temporal decision advantage it provides our elected leaders. Warnings and indications, and assessments like this — they ought to be kept out of the public domain until the decision has been made. (If lawmakers had actually read those fabled Iraq WMD National Intelligence Estimates, they might have protested more than they did about the war; but few bothered to read them, even though they'd been made widely available. So give credit to Lamborn for reading. Still.)
I would not be surprised if the person marking the DIA document meant to classify that line as (C) Confidential or (S) Secret, and merely, perhaps because he or she was rushing to prepare the estimate in light of world circumstances, just glazed over that particular portion marking.
By the way: Every Defense Department declassification guide includes the following warning: