Toward agentic AI: User acceptance of a deeply personalized AI super assistant (AISA)

Abstract

Recent scholarship underscores the transformative potential of generative AI in shaping consumer decision-making, preferences, and overall brand satisfaction. Among these technologies, chatbots and AI voice assistants are increasingly deployed in marketing to influence consumer behavior. A critical question, however, is whether consumers are willing to accept a new generation of such technologies. In July 2025, OpenAI introduced the agent mode of ChatGPT, which represents a shift toward highly personalized, multimodal, and autonomous systems. This study defines these systems as AI super assistants (AISA). Informed by the broader literature on AI adoption and consumer behavior, an adapted AIDUA model with perceived risk is proposed. Survey data from the Philippines (N = 407) was analyzed using combined PLS-SEM and NCA methods. The results show that users appear increasingly confident in their ability to engage with new AI technologies, indicating that they do not feel overwhelmed but instead perceive AISA's new features as manageable. Hedonic motivation, novelty value, performance expectancy, and effort expectancy were identified as necessary conditions for user acceptance, while perceived risk is a necessary condition for objection. These findings offer new insights into user perception toward AISA, with implications for responsible AI design and deployment.

Ähnliche Arbeiten