Read the original article here.

---

Digital Manipulation: How They Use Data to Control You - An Exploration of Computational Propaganda

The digital age has revolutionized how information is shared, but it has also opened sophisticated new avenues for manipulation. In the context of "How They Use Data to Control You," a critical concept to understand is Computational Propaganda. This refers to the use of advanced computing power and data analysis to spread misleading information and influence public opinion or behavior on a massive scale, often without the targets even realizing they are being manipulated.

This resource delves into the nature of computational propaganda, explaining its components, how it leverages data, and why it is a potent tool for digital manipulation.

What is Computational Propaganda?

Computational propaganda is a modern form of propaganda that integrates traditional manipulation techniques with the power of digital technology.

Computational Propaganda: The use of computational tools, primarily algorithms and automation, to distribute misleading information and manipulate public opinion or behavior through social media networks and other digital platforms.

Essentially, it's about using sophisticated software, automated systems, and data analysis to achieve the goals of traditional propaganda – swaying beliefs, attitudes, and actions – but with unprecedented scale, speed, and precision.

The Digital Advantage: Enhancing Traditional Propaganda

Traditional propaganda often relied on mass media (newspapers, radio, television) to broadcast messages designed to appeal to emotions and bypass rational thought. While these methods were effective, they were relatively blunt instruments. Digital technology enhances these established techniques in several key ways:

1. Scale and Speed: Digital tools allow for the instantaneous spread of messages to millions or billions of people worldwide.
2. Automation: Tasks like posting, liking, sharing, and commenting can be automated, creating the appearance of widespread organic activity.
3. Data Analysis: The vast amount of data available online provides deep insights into individuals and groups, allowing for highly targeted messaging.
4. Personalization: Messages can be tailored to exploit specific biases, emotions, and interests of different audiences.
5. Mimicry: Automated accounts (bots) can be created to appear as real people, making the propaganda seem more credible and widespread.

These enhancements transform propaganda from a broadcast mechanism into a highly dynamic, adaptive, and often covert form of manipulation.

The Mechanics: How Data and Tools Enable Control

At the heart of computational propaganda is the strategic combination of data analysis and computational tools. This process can be broken down into several stages:

1. Data Collection and Analysis: Understanding the Target

The foundation of digital manipulation is data. Computational propaganda leverages "big data" collected from various online sources.

Big Data: Extremely large data sets that may be analyzed computationally to reveal patterns, trends, and associations, especially relating to human behavior and interactions.

Sources of this data include:

• Social Media: User profiles, posts, likes, shares, comments, network connections, group memberships, location data, and even the timing of activity reveal interests, political leanings, emotional states, and social vulnerabilities.
• Internet of Things (IoT): While perhaps less direct for political propaganda than social media, data from connected devices (smart speakers, fitness trackers, smart homes) can potentially provide insights into daily routines, lifestyle, and potentially even health or financial status, further building a comprehensive profile of an individual or household.
• Browsing History & Online Activity: Websites visited, search queries, online purchases, and time spent on various platforms offer clues about interests, beliefs, and susceptibility to certain types of information.

This vast ocean of data is analyzed using algorithms to:

• Identify Target Audiences: Pinpoint specific groups or individuals who are most likely to be receptive to, or influenced by, a particular message. This could be based on demographics, interests, political affiliation, psychological traits inferred from language use, or even predicted persuadability.
• Understand Vulnerabilities: Analyze what topics, emotions (fear, anger, hope), or existing biases are most likely to resonate with or provoke a desired reaction from the target audience.
• Map Social Networks: Identify influential users or clusters of users within a network through which propaganda can spread effectively.

2. Computational Tools: Automation and Algorithms

Once the targets and messaging strategies are determined based on data analysis, computational tools are deployed for execution.

Algorithms: A set of rules or instructions to be followed in a computation or other problem-solving operations. In this context, algorithms can be used to analyze data, identify patterns, automate tasks, and determine which content is shown to whom.

Automation: The use of largely automatic equipment in a system of manufacturing or other production process. In digital manipulation, this refers to using software to perform tasks automatically, such as posting, liking, sharing, or creating content.

Key computational tools include:

• Bots (Autonomous Agents): Software applications that run automated tasks (scripts) over the internet. In computational propaganda, bots are used to mimic human behavior on social media.

  Internet Bot: A software application that runs automated tasks (scripts) over the internet, typically performing simple and repetitive structured tasks at a much higher rate than would be possible for a human alone. Malicious or deceptive bots mimic human users to manipulate online environments. Bots can be used to:

  • Amplify Messages: Automatically like, share, or retweet propaganda messages to make them appear more popular or credible than they are ("astroturfing").
  • Generate Content: Post automated messages, comments, or even simple articles.
  • Create Fake Personas: Manage multiple fake accounts with seemingly human profiles (profile picture, bio, posting history) to interact with real users and spread messages.
  • Flood Conversations: Drown out legitimate discussions or counter-narratives with a deluge of automated posts.
• Targeting Algorithms: The same algorithms used for legitimate advertising (like those on social media platforms) can be repurposed to deliver specific propaganda messages only to the identified target audiences. This is known as microtargeting.
• Content Recommendation Algorithms: Understanding how platform algorithms prioritize content allows manipulators to optimize their messages (e.g., using specific keywords, formats, or interaction patterns) to gain maximum visibility in users' feeds, even if the content is false or misleading.

3. Execution: Targeted Distribution and Manipulation

Combining data insights with computational tools allows for the precise delivery of manipulative content. Instead of broadcasting a single message widely, computational propaganda employs microtargeting to send highly specific, often emotionally charged, messages to narrow segments of the population identified through data analysis.

Example Use Case: Imagine a foreign actor wanting to sow discord in a country before an election.

1. Data Analysis: They analyze social media data to identify groups with existing grievances, for example, people concerned about immigration, or those distrustful of mainstream media. They might identify specific emotional triggers associated with these groups (e.g., fear, economic anxiety).
2. Message Crafting: They craft multiple versions of propaganda messages, each tailored to resonate with a specific identified group and exploit their particular emotional triggers and biases. One group might receive messages exaggerating immigrant crime rates, another might receive messages claiming the election is rigged by elites.
3. Tool Deployment: Bots are created to appear as disgruntled citizens belonging to these groups. These bots start posting, liking, and sharing the tailored messages. Paid human trolls might also be employed.
4. Targeted Advertising: Small amounts of money are spent on social media ads using the platform's targeting tools to push these specific messages directly into the feeds of individuals identified in the data analysis phase as susceptible to that particular narrative (e.g., targeting users who have expressed anti-immigrant sentiment or distrust of government online).
5. Amplification: The bots and potentially real users influenced by the initial push amplify the content, making it appear to be spreading organically and gaining traction. This creates a false sense of consensus or widespread belief.

Through this process, different individuals receive different manipulative messages designed specifically to influence them, based on what the data reveals about their individual susceptibility.

The Goal: Control Through Influencing Opinion and Behavior

The ultimate aim of computational propaganda, within the context of "How They Use Data to Control You," is to influence thought, belief, and ultimately, behavior. This control can manifest in various ways:

• Influencing Political Outcomes: Spreading disinformation to sway votes in elections, suppress voter turnout, or create political instability.
• Shaping Public Discourse: Pushing specific narratives, polarizing public opinion on social issues, or distracting from important topics.
• Damaging Reputation: Spreading false information to discredit individuals, organizations, or governments.
• Manipulating Markets: Spreading false rumors about companies or economies to influence stock prices.
• Undermining Trust: Eroding trust in institutions, media, science, or democratic processes.

By leveraging data to understand individuals and using computational tools to deliver precisely targeted, emotionally manipulative messages, computational propaganda seeks to control the informational environment and, consequently, human decision-making on a grand scale.

Pioneering Research in Computational Propaganda

The study and identification of computational propaganda as a distinct phenomenon have been significantly advanced by research teams dedicated to understanding its impact. Pioneering work in analyzing this concept has been conducted by the team of Philip N. Howard at the Oxford Internet Institute. Their research since 2012 has been instrumental in mapping how social media and digital tools are used by political actors and others for manipulation, building on earlier studies of media effects on the public.

Understanding computational propaganda is crucial in the digital age. It highlights how the vast amounts of data we generate online, combined with increasingly sophisticated computational tools, can be weaponized not just to sell us products, but to influence our thoughts, feelings, and actions, fundamentally altering the landscape of communication and control.