Paper Type
Complete
Abstract
Artificial intelligence is playing an increasingly important role in user interaction, with AI agents enabling personalized, emotionally appealing communication. At the same time, some AI agents deliberately employ manipulative strategies to influence user behavior, acting as 'dark agents'. However, it is unclear to what extent people perceive such agents as manipulative and how they react to them. Against this background, Information Manipulation Theory is applied in the context of dark agents to address this research gap. The empirical results show that violations of the communication maxims – quality, quantity, relevance, and clarity – by an AI agent increase perceived manipulation, impairing the development of a parasocial preference for it. In addition, people with low persuasion knowledge react more strongly to perceived manipulation attempts. These findings emphasize the importance of transparent, clear and relevant communication when designing AI agents, and highlight the ethical challenges of manipulative strategies in human-AI interaction.
Paper Number
1900
Recommended Citation
Fröbel, Lara, "Perceived Manipulation by Dark Agents – How Violations of Communication Maxims Affect the Development of Parasocial Preferences" (2026). AMCIS 2026 Proceedings. 5.
https://aisel.aisnet.org/amcis2026/sig_svs/svs/5
Perceived Manipulation by Dark Agents – How Violations of Communication Maxims Affect the Development of Parasocial Preferences
Artificial intelligence is playing an increasingly important role in user interaction, with AI agents enabling personalized, emotionally appealing communication. At the same time, some AI agents deliberately employ manipulative strategies to influence user behavior, acting as 'dark agents'. However, it is unclear to what extent people perceive such agents as manipulative and how they react to them. Against this background, Information Manipulation Theory is applied in the context of dark agents to address this research gap. The empirical results show that violations of the communication maxims – quality, quantity, relevance, and clarity – by an AI agent increase perceived manipulation, impairing the development of a parasocial preference for it. In addition, people with low persuasion knowledge react more strongly to perceived manipulation attempts. These findings emphasize the importance of transparent, clear and relevant communication when designing AI agents, and highlight the ethical challenges of manipulative strategies in human-AI interaction.
Comments
SIG SVS