Social bots are automated social media accounts that generate content or engage in coordinated behavior without direct human intervention. Bots range from benign utility bots (weather alerts, news feeds) to sophisticated propaganda machines used in political manipulation, misinformation amplification, and information warfare.
Bot types and functions
Benign/utility bots: Automated feeds (weather, news headlines), reminder services, conversational bots
Amplification bots: Automatically retweet, like, or share content from operators; used to inflate reach and visibility of political messages
Propaganda/coordination bots: Systematically spread false or misleading narratives; part of organized disinformation campaigns like Internet Research Agency operations
Troll bots: Harass individuals, incite conflict, degrade public discourse
Spam bots: Commercial advertising, phishing, scam promotion
Capabilities and impact
- At scale: Coordinated bot networks can amplify content to millions of users with minimal human oversight
- Strategic timing: Bots target influential accounts and coordinate posting to maximize algorithmic amplification
- Evasion: Sophisticated bots mimic human behavior (account age, follower ratios, posting patterns) to evade detection
- Polarization: Bot-amplified inflammatory content reinforces group-specific narratives and widens political divides
Detection challenges
- Sophistication: Well-resourced operators (state actors, organized crime) engineer accounts that closely mimic humans
- Concept drift: Bot operators continuously adapt to evade detection systems
- Class imbalance: Bots may be rare in some populations, complicating supervised learning
- Privacy/attribution: Detecting bots requires platform data; limited API access restricts research
- False positives vs. false negatives: Removing humans misidentified as bots harms users; missing bots enables disinformation
Key papers and talks in this wiki
- Persily & Tucker (2020) — Social Media and Democracy: The State of the Field and Prospects for Reform — Comprehensive edited collection with dedicated chapter on bots and computational propaganda (Woolley); covers bot-driven amplification, automation tactics, platform affordances enabling coordinated behavior, and policy responses.
- 2018-badawy-russian-interference-twitter — Characterizes bot prevalence among users who amplified Russian trolls; ~5% of both liberal and conservative spreaders were bots; bots present across ideological spectrum but concentrated in conservative amplification chains
- 2015-ferrara-rise-social-bots — Foundational survey of social bot phenomena and detection methods
- BotOrNot: A System to Evaluate Social Bots — Canonical bot detection tool using 1,000+ features
See also
Papers on this topic
Spread & networksPlatforms, policy & society
Beyond "Bots and Trolls" — Understanding Disinformation as Collaborative Work
Spread & networksDetection methods
Analyzing the Digital Traces of Political Manipulation: The 2016 Russian Interference Twitter Campaign
Detection methodsSpread & networks
Social Media and Democracy: The State of the Field and Prospects for Reform