The State of Information Warfare, Part I: The Method

Erik Nolan Americas, Analysis, Conflict & Security, Law & Institutions, Russia & FSU

5 minute read

Information warfare’s role in the strategic environment has always been relevant, but until recently it was relegated as a secondary component in traditional battlespace dimensions (incorporated as the fifth battlespace dimension as early as 1995). In the wake of such events as the Russian involvement in the 2016 American presidential election, Brexit, and China’s continued centralization and ascendancy, active measures are being utilized and refined on a scale not seen since the Cold War—and arguably, are enjoying a much higher success rate than previously thought possible.

Active measures, typically used by state intelligence agencies, are deliberate actions taken to influence or impact an adversary through overt or covert means. Rather than focus on intelligence collection, active measures typically range from operations as simple as gauging the international response of a military action, to complex, decades-long culture wars. Big data’s contribution to active measures today concerns how personal information is weaponized to politically subvert an adversary. Understanding the role of big data in information warfare has now become the concern of governments as well as private enterprise in recent months.

The trail begins at the 2012 U.S. presidential elections. Looking to lock in a second term, Barack Obama’s campaign had to advertise their message of loyalty to very specific sets of voters in order to ensure voter loyalty (staving off defections to Republican candidate Mitt Romney) and encourage active participation in the electoral process (such as registering friends to vote or volunteering for the campaign). Voters in swing states were a top priority as always, but traditional demographic analysis was being supplanted by new, hyper-quantitative statistical modelling to find—down to the most minute personal details—the makeup of each and every single voter that had to be targeted.

The staffers who procured such data, which was harvested in a windowless room at campaign headquarters in Chicago, had initially been relying on set-top box data to analyze viewer trends and thus target advertisements around them, until a cheaper and more refined method soon augmented and superseded television set-top box analysis: Facebook. By getting supporters to login to the campaign via Facebook credentials, analysts for the campaign could harvest and quantify specific personal information—termed psychographics by researchers—into “persuasion” scores, not dissimilar from the metrics of a psychological personality test. The scores informed the campaign of which ads specifically to target individual voters with.

However, the analysts weren’t simply using data through participants of the Facebook login. Psychographics were being actively compiled throughout participants’ entire friend lists using algorithms tailor-made to interface with Facebook’s code by examining a participant’s photos where their friends were tagged. This gave campaign analysts psychographic data for millions of Americans which would then be cross referenced with voter data to more accurately target political ads and devote campaign resources. For its part, Facebook had knowledge of the data mining but did not clamp down on such practices until after Obama secured his second term (according to one former analyst, Facebook tacitly endorsed the campaign in spite of the data mining).

In 2014 and 2015, Cambridge Analytica had set off the site’s alarm bells when it had scooped up millions of profile’s worth of American user data, similar to what had happened in 2012. The terms of service for data collection having changed since the previous mass-collection, Facebook demanded that Cambridge Analytica turn over or destroy the data. Cambridge Analytica said it complied with the request, but emerging reports dispute that claim.

Cambridge Analytica had used Facebook surveys to gauge American’s perceptions about topics like Vladimir Putin and Russian expansionism for a Russian oil company. All in all, the net was cast wide enough to establish psychographic information on 50 million users. The same firm ended up advising the Trump campaign during the 2016 presidential election and has since become the subject of many a government inquiry.

Similar to the 2012 campaign, the main method of psychographic data gathering was conducted through a Facebook login to an application that administered a personality test. And while only 270,000 users gave permission for their data to be collected in the case of Cambridge Analytica, similar data-mining algorithms expanded the data collection into the millions of users. Watching advertisements, taking surveys, “liking” content online—even something as innocuous as a cat video—was and continues to be compiled into psychographic data on an individual level, for every individual and their friends and associates.

The importance of intelligence gathering—no matter how menial the details are—is an essential component of waging information warfare. While intelligence gathering used to be limited to strategic military operations, heads of state and their subordinates, the state of information warfare today mobilizes the entire population to execute an operation. Part II of this series will assess the role of the population in the spate of active measures emanating from Russia over the past few years.