Critics have targeted a recent study on how emotions spread on the popular social network site Facebook, complaining that some 600,000 Facebook users did not know that they were taking part in an experiment. Somewhat more disturbing, the researchers deliberately manipulated users’ feelings to measure an effect called emotional contagion.
Though Cornell University, home to at least one of the researchers, said the study received no external funding, it turns out that the university is currently receiving Defense Department money for some extremely similar-sounding research—the analysis of social network posts for “sentiment,” i.e. how people are feeling, in the hopes of identifying social “tipping points.”
The tipping points in question include “the 2011 Egyptian revolution, the 2011 Russian Duma elections, the 2012 Nigerian fuel subsidy crisis and the 2013 Gazi park protests in Turkey,” according to the website of the Minerva Initiative, a Defense Department social science project.
It’s the sort of work that the U.S. military has been funding for years, most famously via the open-source indicators program, an Intelligence Advanced Research Projects Activity (IARPA) program that looked at Twitter to predict social unrest.
If the idea of the government monitoring and even manipulating you on Facebook gives you a cold, creeping feeling, the bad news is that you can expect the intelligence community to spend a great deal more time and money researching sentiment and relationships via social networks like Facebook. In fact, defense contractors and high-level U.S.intelligence officials say that social network data has become one of the most important tools they use in the collecting intelligence.
Defense One recently caught up with Lt. Gen. Michael Flynn, the director of the Defense Intelligence Agency who said the U.S. military has “completely revamped” the way it collects intelligence around the existence of large, openly available data sources and especially social media like Facebook. “The information that we’re able to extract form social media—it’s giving us insights that frankly we never had before,” he said.
In other words, the head of one of the biggest U.S. military intelligence agencies needs you on Facebook.
“Just over a decade ago, when I was a senior intelligence officer, I spent most of my time in the world of ‘ints’—signals intelligence imagery, human intelligence—and used just a little bit of open-source information to enrich the assessments that we made. Fast forward to 2014 and the explosion of the information environment in just the last few years alone. Open-source now is a place I spend most of my time. The open world of information provides us most of what we need and the ‘ints’ of old, they enrich the assessments that we’re able to make from open-source information.”
Open-source intelligence can take a variety of forms, but among the most voluminous, personal and useful is Facebook and Twitter data. The availability of that sort of information is changing the way that DIA trains intelligence operatives. Long gone are the spooks of old who would fish through trash for clues on targets. Here to stay are the eyes looking through your vacation pictures.
“We train them differently even than we did a year ago because of the types of tools we have. There are adjustments to the trade craft, and that’s due to the amount of information we can now get our hands on,” Flynn said.
The growth of social media has not just changed day-to-day life at agencies like DIA, it’s also given rise to a mini gold rush in defense contracting. The military will be spending an increasing amount of the $50 billion intelligence budget on private contractors to perform open-source intelligence gathering and analysis, according to Flynn. That’s evidenced by the rise in companies eager to provide those services.
Some of them are well known like Palantir, the Silicon Valley data visualization giant that’s been featured prominently in Bloomberg Businessweek and has graced the cover of Forbes. Collecting or analyzing social network data wasn’t something they originally wanted to get into according to Bryant Chung, a Palantir employee. Palantir doesn’t market itself as a data collection company. They provide a tool set to help agencies visualize and share data.
The company worried that partnering with the intelligence community to do social network data collection could hurt their reputation among the tech community, increasingly wary of the government, according to Chung. When the company was approached by NATO and some U.S.intelligence groups, they decided to explore the marketplace for sentiment analysis of social network data.