Social Media Intelligence

U.S. Government employees who serve abroad representing our nation’s interests perform a critical task that also often puts them at great personal risk.  Since World War II, the State Department has had seven ambassadors murdered in the course of their duties.  Serving as an ambassador is more dangerous than serving as a general officer in war time and assassination is not their only threat. In 1979, Iranian revolutionaries seized the entire US embassy in Tehran and held its occupants for 444 days.  That same year Pakistani militants stormed the US embassy in Islamabad, Pakistan and burned it to the ground, killing two embassy personnel in the process. 

Sadly, other examples abound.  The tragic bombing of the U.S. Marine Barracks in Beirut in 1983 has somewhat obscured from memory the equally tragic, though less deadly, bombing of the American embassy in Beirut earlier that year.  That event prompted the Inman report, which called for the creation of the Bureau of Diplomatic Security and the Diplomatic Security Service.  The DSS provides embassies with Regional Security Officers who serve as the ambassador’s primary security advisor and also controls an array of assets including local and Marine Security Guards.  Clearly, in Benghazi the best efforts of these men and women were not enough to protect the consulate and its occupants given the overwhelming nature of the threat and the horribly inadequate security provided by the host nation. 

The families of the deceased, the Foreign Service and civil servants and the people of the United States deserve a full accounting of what happened in Benghazi.  We should let that accounting play out in an objective and thorough fashion, but that does not preclude us from drawing preliminary findings from what happened.

One such finding is that the US diplomatic presence in Libya, and the Benghazi consulate in particular, was horribly lacking in situational awareness.  While there were intonations that something bad was going to happen in Benghazi on September 11, 2012, such as Sean Smith’s forum message that local nation guards appeared to be casing the consulate, there simply wasn’t sufficient signal of a threat to drown out the ubiquitous noise that comes with supporting a diplomatic presence overseas.  For all its funding and assets, the vaunted American intelligence community simply cannot be everywhere at once.  It does, of course, make its best effort to support intelligence consumers, the State Department among them. But even a decade into the war on terrorism, much of its collection is tied to large stove-piped programs modeled on the Cold War world and unsuited to the State Department’s needs including early warning.

According to emails acquired by the Washington Post, an insurgent group claimed responsibility for the consulate attack a few hours afterwards, an important (albeit unverified) clue to the assailants’ identities.  The source of this tidbit was Facebook.  Secretary Clinton quickly pointed out that a Facebook quote is not authoritative, but the idea that insurgents would claim successes on social media sits on firm ground.  Islamic terrorists have been notorious publicity hounds since the early days of the war on terror and their horrific beheading videos.  Insurgents in Syria regularly use social media.  Joe Holliday, an analyst at the Institute for the Study of War stated in a recent talk that Facebook and YouTube were two of his best sources of information on the Free Syrian Army.  Libya has a lower internet penetration level than Syria, but during the 2011 conflict volunteers supporting the United Nations successfully tracked human rights abuses and troop movements through a network of sources using Twitter and Facebook.

Given the importance of social media, even in countries with relatively low “digital indexes,” the State Department should continue to monitor social media on so it can alert embassies and nearby military units if something is afoot. Such a capability might not have prevented the Benghazi attack, but it might do so in the future.

Social media analysis is growing smarter every day.  Many business tools exist to perform sentiment analysis--essentially determining when people are likely to switch banks or whom they would choose to vote off during reality shows.  In addition to “tracking which way the haystack is moving”, some social media tools can find the needle in the haystack as well.  These tools monitor social media on an industrial scale and can provide early warning by alerting watch officers to significant activity patterns or levels on blogs, comment boards, or Twitter.  During Hurricane Sandy, Virtual Operations Support Teams in places like Suffolk County, New York were able to crowd source this function and respond very quickly to requests for information regarding availability of shelter, communications systems, and gasoline for affected victims.  Computer algorithms set to specific parameters for individual countries could do the same for embassies. 

Policy makers should adapt this commercial technology to make greater sense of the signal that abounds in social media channels.  The costs of such systems would be modest when compared to the sums spent on high cost sensors purchased elsewhere in the intelligence community.  Also, social media monitoring is inherently flexible; while drones and their launch/recovery crews could take several days to move from one airfield to another, social media monitoring can quickly pivot from a natural disaster in one country to, say, civil unrest in another.

Of course, an asset such as this will not solve all problems.  Computer algorithms must be trained.  The phrase “Houston rockets” written in a tweet could refer to a terrorist attack or (more likely) a basketball team.  Humans must be in the loop.  Linguists, especially those familiar with slang in the foreign language, would play another key role in this fusion of big data, analysis, and social media.  Finally there is deception.  Social media venues like Twitter are subject to spoofing by false reports.  This is a concern, but not a new issue for intelligence professionals who have a long history of dealing with the lies of double agents and the camouflage of secret bases.  Any social media intelligence system should be tested and hardened against possible spoofing.  In fact, social media may have an inherent stabilizing element in that it “fact checks” itself.  During Hurricane Sandy, a hedge fund analyst named Shashank Trapathi tweeted a series of lies under the Twitter handle @comfortablysmug.  His statement that the New York Stock Exchange had flooded was debunked within a matter of hours by video footage of a dry trading floor.