AI helps with our daily tasks, often influencing our choices, sometimes without us even noticing it. However, when we are aware of AI’s involvement, research shows we usually prefer to make decisions on our own. In their latest article, Patricia Rossi, researcher and associate professor of Marketing at SKEMA Business School, and her co-author, Mariyani Ahmad Husaini from NEOMA Business School, explore the factors that lead us to accept, or reject, AI in our purchasing decisions.
Much of what we do these days seems to bear the mark of artificial intelligence (AI) and algorithms. They are supposed to make our lives easier by taking over certain tasks. But do we want their help to make our purchase choices?
The decision-making process can be complex: identifying what we need or want, gathering information on the different options available, comparing them, and finally making a choice. This process can be more or less lengthy: the more involved we are, the more thought we put into each stage of the process.
The choice of AI
In a purchase, AI can help in at least two ways. First, by recommending something, it narrows down our choices. It takes on the task of gathering information about competing products and comparing alternatives, which can reduce our “choice autonomy.” Second, AI can actively participate in the final purchasing decision, even taking action without our input—taking over our “decision autonomy.” Smart refrigerators, for example, determine when to restock and what to buy. Similarly, the online service Boxed.com ships products based on predictions of when customers will run out of the items.
These types of AI initiatives offer several benefits, such as saving time, reducing mental effort, and potentially saving money. However, we still want to keep our autonomy and make our own decisions. In short, we want to be our own masters! Autonomy has important psychological benefits—it allows us to express our values and shape our identity. Ultimately, autonomy plays a key role in our overall well-being.
Our recent study sought to explore this ambivalence. We developed several experimental studies to understand when autonomy influences the adoption of AI tools in purchasing decisions. Therefore, we isolated choice autonomy and decision autonomy to understand if they individually change our likelihood of adopting AI tools.
Freedom, even when choices are complicated
What do our studies show? Unsurprisingly, the lower the perceived choice autonomy and the perceived decision autonomy, the lower the AI adoption likelihood. In other words, when consumers feel they keep their choice and decision autonomy, they are more likely to adopt AI-powered tools.
But what happens when AI is used to assist with a complex purchasing decision? For instance, when a consumer has to choose based on 20 important attributes instead of just 3. Would they be more willing to give up their autonomy? Initially, we thought so. We assumed that in complex situations, AI could offer valuable support. Factors like limited attention, numerous product attributes, and increased cognitive effort typically lead us to rely more on algorithm-based recommendations. However, to our surprise, even in these complex scenarios, consumers prefer to retain their choice and decision autonomy.
Read also: AI: Apocalypse or Renaissance for Artists?
Overall, the results of our research show that customers’ desire to retain their autonomy outweighs the need to reduce the time and effort involved in making complex decisions. We do not want AI to completely take over our purchasing tasks, because our autonomy matters.
When identity is in play
Does it mean we always prefer to keep our choice autonomy and decision autonomy instead of handing them to AI? In most situations, yes. Except… when AI can help us buy things required to perform activities we consider important to our identity. Fishing, baking, and running, for example, are identity-relevant activities. Yet previous studies have shown that when technology takes over identity-relevant tasks, this leads to technology aversion. Not surprisingly, we want to perform those activities ourselves. After all, we cannot consider ourselves bakers when a machine does most of the work for us!
Instead, our research shows that when an activity is important to us, we are happy to relinquish our purchasing autonomy to AI if it ends up helping us. For example, an avid runner is more likely to let an AI tool buy their running shoes than an occasional runner.
These effects occur because AI-powered shopping tools help consumers achieve their identity-related goals, while allowing them to take ownership of the outcome. If, for example, the avid runner lets the AI take care of the shoe buying, they will have more time and energy to spend on running. They can then unashamedly claim: “I let AI buy my running shoes, so I could spend more time doing what really matters to me: running!”
This article was republished from The Conversation under a Creative Commons license. Read the original article in French
