How automated accounts mimic human behavior to manipulate public opinion
In short: Social bots are no longer just simple scripts. Using artificial intelligence and large language models, these automated accounts can mimic human dialogue and coordinate messaging to create a false impression of organic grassroots support, potentially influencing everything from social media trends to election outcomes and stock prices.
Social bots are fully or partially automated accounts designed to replicate regular user actions, such as posting content, liking, or chatting. While some include a human-in-the-loop, many use artificial intelligence to perform social media actions. These bots can operate individually or in coordinated networks to engage in inauthentic behavior, often used to spread disinformation or commit ad fraud by artificially boosting engagement metrics.
The motivations behind these bots range from commercial to political. Corporations may use automated agents for customer service, which has contributed to layoffs in human customer-service roles. More maliciously, bots can be used to inflate a user's apparent popularity through fake likes, reposts, or followers. This manufactured influence can be bought and sold on a black market, allowing brands or organizations to create a deceptive sense of massive online presence.
Detecting these accounts is a significant technical challenge. Traditional manual identification is inefficient and costly. Modern research focuses on behavioral feature analysis to distinguish bots from humans. One approach uses a Random Forest model, which has shown an accuracy rate approximately 5% higher than decision tree or neural network models. Another advanced method, the ESA-BotRGCN framework, utilizes emoji-driven sentiment analysis and Relational Graph Convolutional Networks. When tested on the TwiBot-20 benchmark dataset, this multi-modal approach achieved a superior accuracy of 87.46%, demonstrating that analyzing emotional cues and social topology can effectively identify malicious bots.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Research on social bot identification through behavioral feature analysis.
- Also: Social bot
- Also: Emoji-Driven Sentiment Analysis for Social Bot Detection with Relational Graph Convolutional Networks.
To cite this card, link to https://intellectplus.co.uk/app/health-social-bot.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Cybersecurity · More in Technology · Latest · A–Z · All topics
