Keeping Up with the Policymakers: The Unclassified Tearline


Without wading into the politics of the imbroglio over former Secretary Hillary Clinton’s emails, an important yet overlooked point is that U.S. policymakers rely heavily on unclassified email systems to conduct their daily business. According to an article by Steven Myers of The New York Times, a review of policymaker email practices revealed that officials throughout the Obama administration routinely exchanged sensitive information — whether genuinely classified or not remains hotly debated, regardless of the FBI’s recent decision — over unclassified networks. The practice may not be exclusive to the Obama team, either. When confronted with allegations that some of his sensitive emails had been post-marked classified, President George W. Bush’s Secretary of State Colin Powell exclaimed that “we might as well shut the department down” if ambassadors are not allowed to provide their private insights to the secretary over unclassified email.

Rightly or wrongly, routine use of unclassified email systems to share sensitive (but hopefully not classified) information is just the tip of the iceberg when it comes to how policymakers seek to leverage the tremendous insights they receive from their staffs and the U.S. intelligence community. In the particular case of the intelligence community, the reality is that while most policymakers greatly appreciate classified intelligence assessments, the utility of those assessments is limited unless intelligence agencies can keep up with policymakers’ operational tempo and adapt their work to the policymaker’s work environment. Thus, the intelligence community should consider remodeling its intelligence assessments to fit the new paradigm. Specifically, I argue here that whenever possible, the intelligence community should append unclassified assessments to its classified products — let’s call this the “unclassified tearline.”

The creation of tearlines — or classified information downgraded from a higher to lower level of classification, such as from Top Secret to Secret — are already commonplace in the intelligence community. Entirely unclassified tearlines, however, are unexplored terrain. Unclassified tearlines could convey the bottom line and potential implications of the classified story in unclassified terms, while at the same time obscuring sensitive sources and methods. Unclassified tearlines would be non-attributable to individual agencies to mitigate sensitivities of public disclosure and potential ridicule for being wrong. Moreover, unclassified tearlines would not be produced on every topic — in fact, far from it. Only the least sensitive regional and functional topics would be covered, and as agencies became more comfortable with the concept, they could produce higher volumes of unclassified analysis at their own pace.

In truth, the intelligence community already has a number of unclassified intelligence initiatives. For example, the Office of Naval Intelligence (ONI) has published unclassified assessments on both the Russian and Chinese navies, the National Geospatial-Intelligence Agency (NGA) has its Pathfinder initiative that seeks to leverage openly available and commercial instead of classified imagery, and the Open Source Enterprise (OSE, formerly Open Source Center) produces unclassified analysis based on media reporting. Despite these individual efforts, there is no dedicated or institutionalized process requiring unclassified analysis to be produced when feasible.

There are a number of good reasons to build unclassified tearlines into the business process and culture of intelligence agencies. First, policymakers are dealing with “the most diverse array of challenges” to national security ever witnessed, according to recent Congressional testimony by Director of National Intelligence (DNI) James Clapper. Therefore, policymakers frequently encounter an increasingly large and complex number of time-sensitive crises that require the quick exchange of information and ideas. As detailed above, the most effective way of doing so is over unclassified email because the sender does not have to worry about whether the recipient has ready access to classified networks. Unfortunately, this practice does not fit well with intelligence assessments because they are typically classified.  As a result, policymakers tend to rely on their own best judgment in “talking around” the material on the open Internet. This increases the risk of inadvertent disclosure of classified information. Unclassified tearlines would draw a clear line for policymakers regarding what is truly classified and what is not, reducing the chances of accidental leaks. The routinized production of unclassified analysis would also dramatically enhance the usability of intelligence community assessments in these unclassified policymaker communications channels.

Second, apart from the most senior officials, all policymakers struggle to read classified analysis while traveling. According to one former NSC senior director, policymakers when overseas “can only see more sensitive products at certain times and places.” Consequently, “there may hardly be time to scan the material.” Most policymakers are forced to rely on obtaining access to U.S. embassies or military installations to receive intelligence updates. This is usually a significant logistical inconvenience and often results in critical intelligence not getting to its intended recipient. Unclassified tearlines would offer policymakers an ability to view some intelligence analysis in real-time on their mobile devices, enhancing their preparedness for foreign engagements.

Third, policymakers tend to have wide-ranging professional networks to support their efforts. These networks include contacts both inside and outside government. Inside government, policymakers could leverage unclassified analysis to more effectively incorporate non-defense departments and agencies into their policy planning processes. The Departments of Agriculture, Commerce and Health and Human Services, along with USAID, are good examples of government organizations that at times may have a stake in defense matters. Smoother intelligence sharing would also be possible among other government organizations that are non-Title 10 or non-Title 50 — i.e., those without national defense and intelligence collection authorities respectively — such as the Departments of Justice and Homeland Security.

Outside of government channels, policymakers frequently liaise with academics, contractors, consultants, close friends, NGO representatives, think tankers, and former U.S. officials, among others. Policymakers typically seek to either incorporate these individuals into the policymaking process or seek a second opinion from a trusted confidant. Unclassified assessments would allow for easier sharing of intelligence that can support such activities. In the particular case of close confidants, having unclassified tearlines would allow for brainstorming and a reciprocity of views that are not always possible under the current structure. Third party perspectives are essential to combating groupthink and other cognitive biases.

Also outside of government, policymakers would be newly empowered to share unclassified assessments with foreign governments. To be sure, under the current guidance, policymakers are allowed to share intelligence assessments with foreign governments if “consistent with U.S. law and clearly in the national interest to do so.” However, the foreign disclosure process can be slow and burdensome. With unclassified tearlines in hand, the hard work would already have been frontloaded for the policymaker, enabling not only instant disclosure (verbal communication) but also instant release (written communication for retention) of unclassified information to a foreign government. A freer exchange of intelligence might promote closer bilateral ties between the United States and its partners and allies.

Finally, unclassified tearlines would allow policymakers to seamlessly integrate intelligence assessments into briefing books, memos, talking points, and Congressional testimonies. These policy documents are the bread and butter of the policymaking process, yet intelligence does not always factor into them because of security constraints. Like the foreign disclosure process, the downgrading or declassification process is onerous and time-consuming, with mixed results. In one particularly egregious case in the last few years, an intelligence briefer, over the course of several weeks, was unable to receive unclassified talking points from the intelligence community, forcing his Cabinet-level official to quickly jot down notes on a napkin prior to an important bilateral exchange.

If the idea gets any traction in the intelligence community, then there are sure to be detractors that seek to prevent change in the system. Defenders of the intelligence community’s ingrained culture of secrecy will object to disseminating any information that could jeopardize current and future sources and methods. Critics might argue that the analytic production process is already slow enough and would suffer from any added layer of bureaucracy. Others could opine that analysts, knowing in advance that some of their assessments would now essentially be unclassified, might pull their punches and hedge their analysis to prevent ridicule if proven wrong. Another reasonable concern is that the routine disclosure of U.S. intelligence assessments could inadvertently support an adversary’s objectives while undermining the U.S. decision-making advantage.

These are all certainly important counter-arguments that intelligence community leadership would have to address going forward. In general, the tearlines would be kept at a generic enough level to avoid any potential compromise of sensitive sources and methods. They will be thoroughly vetted to ensure this type of information is kept out, in effect obscuring the sources. Only the least sensitive assessments would result in unclassified tearlines. Truly classified stories must remain classified. While it is true that unclassified tearlines would add another layer of review to an already bloated analytic review process, the benefits of assisting policymakers to do their jobs is worth the extra upfront investment. In addition, this initiative would provide a much-needed boost to the intelligence community’s downgrading and declassification capabilities, which have not kept pace with skyrocketing rates of classified products over the last decade., Analysts need not worry about their identities being exposed by leaks of unclassified analysis. The intelligence community has already sufficiently masked analyst identities in the vast majority of its products following unauthorized disclosures. Individual agencies, however, might still worry about their reputations, but keeping the authoring agency’s name anonymous should mitigate this problem.

Separately, the routine disclosure of unclassified intelligence analysis would not hurt America’s decision-making advantage. Unclassified tearlines would only present synopses of lengthier, more nuanced narratives that would remain in the classified sphere. Therefore, U.S. policymakers would still rely on classified assessments for critical decision-making. The purpose of unclassified tearlines would be to give policymakers the opportunity to more easily use and share the highlights of a classified assessment, not to share the entire classified assessment itself. Finally, while adversaries would indeed be keenly interested in acquiring unclassified tearlines, it would be imperative for U.S. policymakers to use their best judgment in sharing these assessments. Inevitably, unclassified tearlines will fall into the wrong hands, but the alternative to an unclassified tearline system is an even more precarious situation in which policymakers continue to try to talk around classified assessments — with their errors resulting in potential harm to national security.

A mid-level policymaker recently told me that classification restrictions impacted his work on a daily basis, causing him to “give up” on the intelligence community for unclassified and sharable analysis. Instead, he felt forced to rely on his own judgment for what was classified and usually looked to non-intelligence open source information as an easy alternative. The intelligence community should strive not to lose intelligence consumers in this way, and if unclassified tearlines were implemented as I suggest here, they would be less likely to.


Derek Grossman is a senior project associate at the nonprofit, nonpartisan RAND Corporation. He previously served at the Defense Intelligence Agency as the daily intelligence briefer to the Assistant Secretary of Defense for Asian and Pacific Security Affairs at the Pentagon. Grossman was the winner of the 2014 “Galileo Competition” which fosters new and innovative ideas in the intelligence community. The views expressed in this article are those of the author and do not reflect the official policy or position of the Defense Intelligence Agency, Department of Defense, or the U.S. government.

Image: U.S. Air Force