Key points are not available for this paper at this time.
Abstract Agentic AI represents a fundamental shift from reactive AI systems (such as LLM-based chatbots) toward autonomous, goal-directed agents capable of decision-making, coordination, and task execution. This technological evolution is expected to reshape how individuals organize daily life, work, and information practices. At the same time, the rise of agentic AI raises important cultural, political, and ethical questions regarding human oversight and the redistribution of agency. This study provides early insights on how consumers anticipate delegating tasks to agentic AI and identifies heterogeneity in these expectations across user segments. Survey data from 407 experienced AI users in the Philippines were analyzed using latent class analysis (LCA) across 12 everyday and work-related activities. A four-class solution emerged, revealing segments of Creative Task Delegators (36.4%), Everyday Convenience Seekers (6.6%), Socially Oriented AI Agent Users (19.2%), and Web-Browsing Automators (37.9%). These distinct profiles illustrate differing orientations toward autonomy, trust, and cognitive offloading. The membership model highlights that age, socioeconomic status, social influence, and hedonic motivation are key factors shaping class membership, indicating that both socio-demographic characteristics and motivational orientations influence how individuals envision delegating tasks to autonomous AI agents. Overall, the findings of this study contribute to ongoing debates about responsible system design, governance, and the societal implications of increasing AI autonomy.
Marc Hasselwander (Sat,) studied this question.