Date
Publisher
arXiv
Generative social agents (GSAs) use artificial intelligence to autonomously communicate with human users in a natural and adaptive manner. Currently, there is a lack of theorizing regarding interactions with GSAs, and likewise, few guidelines exist for studying how they influence user attitudes and behaviors. Consequently, we propose the Knowledge-based Persuasion Model (KPM) as a novel theoretical framework. According to the KPM, a GSA's self, user, and context-related knowledge drives its persuasive behavior, which in turn shapes the attitudes and behaviors of a responding human user. By synthesizing existing research, the model offers a structured approach to studying interactions with GSAs, supporting the development of agents that motivate rather than manipulate humans. Accordingly, the KPM encourages the integration of responsible GSAs that adhere to social norms and ethical standards with the goal of increasing user wellbeing. Implications of the KPM for research and application domains such as healthcare and education are discussed.
Study at a glance
- Question: How does a generative AI agent's knowledge shape its persuasive behavior and influence human user responses?
- Setting and data: Online experiment with 113 participants across fitness, nutrition, and investment scenarios.
- Method: Conceptual framework development with an online experiment and qualitative conversation analysis.
Why it matters
Generative social agents have become highly persuasive communicators, but there is a lack of theorizing and few guidelines to understand how they influence human user attitudes and behaviors.
Key findings
- Providing agents with user-knowledge and context-knowledge increased persuasion effectiveness, primarily by improving user attitudes toward the agent and subsequently toward the persuasive message.
- Qualitative analyses revealed that user-knowledge led to more personalized arguments, whereas context-knowledge enabled more context-sensitive and factually accurate recommendations.
- Manipulating an agent's self-knowledge through personality prompts did not produce significant effects on persuasion effectiveness.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
