Precision-Guided Marketing Frames the Future of Information Operations
3rd Place in The Cyber Edge 2021 Writing Contest
A military-age male left home and traveled through the city, unaware he was being surveilled. Those watching him knew his patterns and preferences. They collected his point of departure, route and destination to predict when he would be most vulnerable for attack. Arriving at a marketplace, he meandered through a few high-traffic areas. Passing down a quiet corridor, he finally provided a clear shot. His smartphone buzzed and its screen flashed: “Two-for-one sale at the nearby pretzel shop!” He was struck by a precision-guided advertisement.
Every day, consumers are targeted with precision-guided marketing from both legitimate sources and covert operators. Backend collectors harvest and siphon digital traces streaming from mobile devices as a precious commodity. These companies monetize and sell personal Internet browsing history, app data, contacts and location services to those intent on influencing behaviors.
Those influencers use recommender systems, which are automated information filtering systems to control content in news feeds, social media environments and streaming media services. These systems know what information—good or bad—will capture a person’s attention. Then the system can deliver that content.
The same commercialized recommender systems that now strongly influence consumers’ decisions are being harnessed by political operatives, governments and other influencers to conduct large-scale information operations on U.S. citizens. Covert information operators’ adoption and adaptation of commercial precision-guided marketing technologies are being used to manipulate unsuspecting citizens and present a growing national security risk. These threats require immediate attention from those in the cybersecurity arena charged with protecting and defending the nation’s information and digital commons.
Information operations seek to influence people’s understanding and behavior. Often associated with deception and other unsavory tactics, at its heart, information operations aspires to instill a set of beliefs even though the affected party is unaware of the influence being exerted.
Decades ago, subliminal messages were conveyed via television. Today, the avenues for influence are multifaceted and far more developed. Audiences are overwhelmed with choice in digital ecosystems. Distracted masses have increasingly become reliant upon automated systems to deliver customized content. Information operators are happy to exploit these systems.
In the early 1930s, Aldous Huxley expressed his distress over propaganda and influence campaigns leveraging mass media’s ability to distract through newspapers, magazines, radio, television and the cinema. Huxley’s concerns were prescient and perhaps understated.
In the 1960s, Herbert Simon agreed with Huxley and predicted that markets would arise to capitalize on the decreasing supply of consumers’ attention in what he labeled as the attention economy. He believed markets would form where commercial organizations and others with demands for attention would pay for its supply.
Today, people increasingly spend more of their attention online. A 2019 Pew Research Study found that 30 percent of American adults report being online “almost constantly.” This finding is approximately 10 percent greater than in 2015. The same study reported that 92 percent of mobile device users reported daily online activity. These findings help explain why information operators have adopted tactics proven to be successful in the attention economy. They have followed their targets to their digital habitats.
In the 1950s and 1960s, the attention economy gave rise to new advertising techniques to attract audiences and keep them engaged. Today, social media platforms and search engines have also evolved their methods. These online marketers now dominate the competition to attract consumers’ attention. Facebook and Google, two of the most powerful information conveyors, now own 60 percent to 70 percent of the annual $80 billion ad-selling markets in the United States.
In The Age of Surveillance Capitalism, Shoshana Zuboff documented the economic relationships and incentives of companies like Facebook and Google that depend upon directing audiences’ attention to consumer products. Their ad revenue comes from auctioning behavioral predictions. Surveillance capitalists use residual data such as search terms, sites visited, likes, shares and location services to catalog individuals and develop models that use this information to predict consumer behaviors, and then sell this information. The consumer thinks the service is free, but in fact, the consumer is the product. Marketers employ these behavioral predictions to deliver precision-guided advertisements.
Precision-guided marketing determines the movies in people’s streaming media queue, the products in their online shopping sites, as well as the headlines in their news feeds. Recommender systems do not necessarily discriminate commercial from noncommercial sources. Information operators have found ways to leverage recommender systems to find and target individuals they believe are susceptible to influence and manipulation.
Although some consumers still obtain their daily news through traditional media sources, within the United States, Pew Research recently reported that perhaps as many as 55 percent of adults consume their daily news via social media. News reporting helps shape audiences’ perspectives on important events and narratives.
Traditional media outlets maintain editorial supervision and can be held liable if they spread falsehoods. Within social media environments, recommender systems do not have oversight and often reflect an individual’s biases.
Myside bias, a form of confirmation bias, leads individuals to engage with content they prefer and messages they like over content challenging their beliefs. Researchers have observed myside bias in how people assess political information sources and share political messages online. Recommender systems appear to be amplifying the effects of myside bias and generating political echo chambers where groups of people with shared views congregate online.
Within echo chambers, stories and content that align with a group’s ideology gains traction and increased shares, often by automated news and content recommender systems. Moreover, automated recommender systems appear to play a disproportionate role in the proliferation of low credibility information such as fake news. These echo chambers provide a space for information operators to weaponize recommender systems to increase social polarization.
Polarization divides societies into distinct groups that share opposite ends on a particular spectrum. It can lead to increased distrust between groups, hatred and civil unrest. Information operators often gain from increasing polarization. Unfortunately, as more people turn to social media news feeds, the more they will be exposed to content tailored by information operators to prey on their biases.
When used appropriately, recommender systems help people navigate the overwhelming abundance of online information. They also enable organizations to employ their advertising resources efficiently. If recommender systems are eliminated or restrictions are imposed that may affect their accuracy, consumers and commercial entities may experience some increased efficiency costs.
Furthermore, implementing policies that may restrict citizens’ access to and distribution of information could approach censorship. In the United States, freedom of speech forms a cornerstone of the democracy. In free and open societies, the clash of ideas provides a method by which citizens can measure competing arguments. In the United States, citizens defend the rights of everyone to voice opinions, even when they venture into distasteful areas. Therefore, heavy-handed censorship of recommender systems is not advised.
The concern here is not about eliminating recommender systems or restricting their legitimate use by lawful citizens. The question is how to best protect unwitting citizens from being exposed to nefarious agents’ use of these systems to conduct harmful information operations. Policy interventions to protect citizens from information operations should focus on the misuse of recommender systems by foreign agents, and the U.S. cybersecurity community should assist.
U.S. cybersecurity authorities protect and defend secure access to networks and information for U.S. citizens. They should seek to protect the information that resides and passes over these networks because these professionals have the technical expertise to detect and deny those entering U.S. cyberspace to conduct information operations. Therefore, they should play a critical role in any broad national security initiative. However, for these efforts to succeed, private organizations operating recommender systems must be involved in helping deny, degrade and defeat foreign operators.
Digital platforms that manage online information streams benefit from a 1996 law, Section 230 of the Communications Decency Act. Through this law, they remain immune from liabilities related to third-party content. As a result, a tragedy of the digital commons now exists, where no one is held accountable for the social costs associated with unregulated digital ecosystems. While some platforms may self-regulate, they deserve guidance from the U.S. and other governments charged with protecting their citizens’ information security.
Digital ecosystems have essentially become polluted with misinformation and disinformation amplified by precision-guided marketing tools. A variety of policy options exist to reduce this pollution. Pigouvian taxes could fine platforms that fail to limit harmful information operations. Additionally, the government could provide subsidies to help pay for information operation mitigation efforts. However, a public-private partnership with increased cooperation to share intelligence on information operations may prove more effective and economically efficient.
National security is a public good required to assure the free and open exchange of all forms of markets, be they economic or otherwise. As such, defending digital ecosystems from information pollution data operators should be a mission the U.S. national security apparatus conducts, particularly cybersecurity agencies, regardless of the actions private companies take. This is not to say that the government should involve itself in arbitrating public debates over fake news or fact-checking every political statement. It can, however, help identify when online personas misrepresent themselves and conduct coordinated information operations online.
The economic incentives that spawned recommender systems are here to stay, and information operators will continue to find ways to exploit them. The time has come for U.S. cybersecurity teams to partner with industry to help preserve the digital commons from information pollution caused by foreign agents.
Lt. Col. Ryan Kenny, USA, created an online forum to foster discussions on emerging technologies at www.militarycommunicators.org. The views expressed here are his alone and do not represent the views and opinions of the Defense Department, U.S. Army or other organizations with which he has had an affiliation.