Unleashing the Machine : Exploring the Dual Power of AI's Agentic Behavior and Anthropomorphic Design in an Agents Decision-Making

Artificial intelligence (AI) use is transforming organizational decision-making. Much of the focus of existing decision-making research has been on AI models' ability to process vast datasets, automate routine tasks, and allow individuals to focus on higher-order thinking [1, 2]. This focus has primarily drawn on the bounded rationality model, which implies that human decision-makers operate within limits of cognitive capacity, time, and accessible information, often leading to satisficing rather than optimal decisions [3]. AI agents, though, unlike traditional decision-making tools, have two added unique characteristics: Agentic Behavior and Anthropomorphic presentation [4]. Agentic behavior allows the shaping of experiences and outcomes. The rapid progress in these AI-based agents presents a risk where they are likely to drive agentic behavior themselves [5]. In addition, the anthropomorphic representation of these agents is likely going to lead to changes in how individuals interact with these agents [4]. This research investigates these two characteristics as they are juxtaposed with the decision-makers' agentic behavior. Drawing from the research on agency theory as well as micro-structuration principles, this research aims to present a socio-technical perspective on employee decision-making. Agency Theory explains the relationship between principals (e.g., employers) and agents (e.g., employees or systems), where agents act on behalf of principals but may have misaligned interests, leading to potential conflicts [6]. In the context of AI, it functions as an agent for employees, filtering, analyzing, and presenting relevant information. The anthropomorphic presentation is defined as the tendency to imbue the real or imagined behavior of nonhuman agents with human-like characteristics, motivations, intentions, or emotions [7], including constant maturity. Researchers have argued that the interaction between humans and AI is likely to lead to a new agency, referred to as synthetic agency [8-10]. The key contribution of this research would be to explain these interactions and how the variance in interactions can influence decision-making in an organization.