Would you trust an AI to buy for you more on a cheap, returnable item — or only once you know it won't switch brands behind your back?

This explores whether the stakes of a purchase, meaning its price and whether you can undo a bad buy, shape how much shopping people will hand over to an AI agent. The honest answer first: the collection has nothing that tests return policies directly, and very little that varies product cost as the main factor. What it does have points somewhere unexpected. Willingness to delegate seems to depend less on the price tag and more on whether protections and control are in place.

The clearest sign is Accenture's survey of more than 25,000 consumers Do consumers actually want to delegate shopping decisions to AI?. Three-quarters say they would trust an agent with purchases, but more than half would still name the brand, and many loyal customers wouldn't let the agent switch them. The authors frame delegation as conditional on trust protections coming first. Return policies are the most obvious real-world example of such a protection, though the survey doesn't single them out. In the same direction, five studies on high-autonomy shopping assistants found that what drives acceptance is the benefit people expect to get, not how in control they feel. That effect grows when goods are scarce and the assistant's help is worth more What drives shoppers to accept autonomous AI assistants?. One reasonable reading is that people weigh stakes against payoff, not stakes alone.

The most direct evidence on price cuts against intuition. In a 484-person Alexa experiment, shoppers who compared fewer options were *more* satisfied and *more* willing to let the voice assistant decide. That held across brand types and price levels Does limiting product choices increase trust in voice assistants?. So in that setting, a higher price didn't make people pull back. Bilgihan and colleagues suggest why the 'how much will I delegate?' question may be too static anyway. Agents that actually complete purchases shape what you want through repeated back-and-forth, so willingness to delegate is something that develops over time, not a fixed threshold Do AI shopping agents reshape what consumers actually want?.

The less obvious angle is that cost changes the *agent's* behavior, not just the shopper's comfort. When order size grew from 1 to 250 bags, one Gemini model stopped comparing unit prices and simply chased the lowest total, and its rate of picking the best option fell from 80% to 35% Does higher purchase volume make AI shoppers search less carefully?. Agents also overreact to price and rating cues far more than humans do Do shopping agents respond more strongly to price and rating cues?. That means the big-ticket purchases where shoppers most want a safety net may be exactly where agents become more rigid and easier to steer. Generous returns would partly be cover for the agent's mistakes, not just the buyer's second thoughts.

If you want to dig into the money side, there's also evidence that sellers win back about a third of an assistant's savings by raising headline prices while cutting add-on fees Do sellers recapture savings when AI assistants help buyers?. That's a reminder that the total cost of delegating includes how the market adapts to it. Return policies as a delegation lever remain an open gap in this collection.