From consumer sovereignty to algorithm-dependency? The case of shopping agents

Abstract

Most consumption and investment decisions are still made by humans today. But we already witness that such decisions are increasingly transferred to algorithms. The newest example, which caught a lot of attention, is the shopping agent “Operator”, which was launched by Open AI in the beginning of 2025. Advanced algorithms not only automatically identify needs but also search for optimal products and purchasing options and possess the potential to carry out the corresponding transactions themselves. They become “autonomous consumer buying agents”, or, in short, shopping agents, which are supposed to enable better consumption decisions in an environment of an ever-increasing and constantly changing variety of products. Consumers may save a lot of time, as shopping agents can screen a much larger number of products much faster than humans. Moreover, shopping agents can evaluate all offers based on criteria, which are supposed to be relevant to individual users. But at the same time, consumers gradually become dependent on shopping agents and lose their sovereignty. They have to “trust” decisions made by AI-based algorithms, which, however, do not necessarily make decisions in favor of consumers but rather in favor of their operators.