BotSim: Mitigating The Formation Of Conspiratorial Societies with Useful Bots

TL;DR

BotSim presents an agent-based model on a Small World network to study conspiracy formation under malicious Bad Bots and interventions by Info-Correction Bots and Good Bots. The framework captures information generation, consumption, and propagation with asymmetric propagation rules and a state-change threshold $t=72$ for humans, enabling analysis of convergence toward a conspiratorial state. Results show Bad Bots alone reliably drive conspiracy formation, while Good Bots—especially at higher deployment—and to a lesser extent Info-Correction Bots, significantly delay or prevent the majority of Bad Humans, with Good Bots offering more robust, scalable protection. The work demonstrates that automated, proactive bot-based messaging can be more resource-efficient than reactive fact-checking, and provides a foundation for designing cyber-social defense strategies.

Abstract

Societies can become a conspiratorial society where there is a majority of humans that believe, and therefore spread, conspiracy theories. Artificial intelligence gave rise to social media bots that can spread conspiracies in an automated fashion. Currently, organizations combat the spread of conspiracies through manual fact-checking processes and the dissemination of counter-narratives. However, the effects of harnessing the same automation to create useful bots are not well explored. To address this, we create BotSim, an Agent-Based Model of a society in which useful bots are introduced into a small world network. These useful bots are: Info-Correction Bots, which correct bad information into good, and Good Bots, which put out good messaging. The simulated agents interact through generating, consuming and propagating information. Our results show that, left unchecked, Bad Bots can create a conspiratorial society, and this can be mitigated by either Info-Correction Bots or Good Bots; however, Good Bots are more efficient and sustainable than Info-Correction Bots . Proactive good messaging is more resource-effective than reactive information correction. With our observations, we expand the concept of bots as a malicious social media agent towards automated social media agent that can be used for both good and bad purposes. These results have implications for designing communication strategies to maintain a healthy social cyber ecosystem.

Figures (8)

• Figure 1: Setup of the simulation model. In the initial setup, Humans are visualized with orange persons, and Bad Bots as red squares. In final setup, Bad Humans are colored black, and the green/red circles represent whether the current set of information consumption queue is good/bad.
• Figure 2: Flowchart of Simulation Logic
• Figure 3: Mean time to Bad Humans Majority and All Bad Humans from singularly varying the proportion of Bad Bots, Info-Correction Bots and Good Bots.
• Figure 4: Response Surface Analysis for varying Bad Bots and Good Bots vs Bad Bots and Info-Correction Bots. The $z$-axis represents time to Bad Human Majority in simulation ticks. The Info-Correction surface exhibits strong concavity ($2\beta_5=-18.6$), indicating diminishing returns with more Info-Correction Bots. The Good Bot surface is almost linear ($2\beta_5=+0.11$), indicating increasing benefits with increased number of Good Bots.
• Figure 5: Fitted defender efficiency functions for good bots and info-correction bots. Here, $T(b,d)$ represents the time to bad-human majority, $b$ is the proportion of bad bots, and $d$ is the proportion of defender bots (either good or info-correction). Coefficients are estimated from quadratic response surface regressions of data from Experiments 4 and 5.