The Role of AI Agents in Autonomous Online Purchasing

Abstract

The rapid development of artificial intelligence is transforming electronic commerce from a model in which consumers actively search, compare, and purchase products toward an emerging model of agentic commerce, in which AI agents can perform substantial portions of the purchasing process on behalf of consumers. AI agents differ from conventional recommendation systems and chatbots because they can interpret goals, plan tasks, interact with digital platforms, retrieve information, compare alternatives, and potentially execute transactions. This development has important implications for consumer behaviour, digital marketing, retail strategy, competition, transaction costs, privacy, trust, and consumer autonomy.

This paper examines the role of AI agents in autonomous online purchasing from a Commerce perspective. It discusses the evolution from traditional e-commerce and recommender systems toward conversational and agentic purchasing. The paper examines how AI agents can identify consumer requirements, search products, compare prices, evaluate reviews, monitor discounts, make recommendations, complete transactions, and provide post-purchase assistance. The analysis suggests that AI agents can substantially reduce information-search costs and decision complexity while increasing convenience and personalization. They may also benefit businesses by improving customer targeting, reducing transaction costs, increasing conversion opportunities, and creating new forms of digital customer relationships.

At the same time, autonomous purchasing introduces substantial risks. Consumers may lose direct visibility into how alternatives are selected, while AI agents may prioritize commercially sponsored products or make decisions based on incomplete or inaccurate information. Privacy concerns may intensify because agents require access to detailed consumer preferences, purchase histories, payment information, and potentially other digital services. Security risks also increase when AI agents possess the ability to perform transactions. Furthermore, autonomous purchasing raises questions regarding consumer responsibility, algorithmic accountability, competition among retailers, and the balance between convenience and consumer autonomy.

The paper concludes that AI agents are likely to become an important component of future digital commerce, but their development should be guided by principles of transparency, consumer control, security, fairness, explainability, and human oversight. Rather than completely replacing consumer decision-making, responsible AI agents should function as delegated decision-support systems that act according to clearly defined consumer preferences and constraints. The future of autonomous online purchasing will therefore depend not only on technological capability but also on trust, governance, market structure, and the ability of consumers to maintain meaningful control over delegated purchasing decisions.