We tested AI shopping chatbots. This is what we found.

Summary

As big-box retailers lean into OpenAI-powered shopping assistants, the Star Tribune ran a hands-on test of AI chatbots positioned as “gift finders” and product discovery tools. The big takeaway: the bots can be genuinely helpful at narrowing choices quickly, but they’re still inconsistent—sometimes overconfident, sometimes vague, and occasionally steering users toward generic bestsellers rather than truly personalized picks. The piece also spotlights the tradeoff retailers are implicitly asking customers to make: better recommendations in exchange for more behavioral and preference data, at a moment when shoppers are increasingly privacy-sensitive. The deeper story is competitive: if AI becomes the front door to e-commerce, whoever controls the agent layer (and the underlying data partnerships) could reshape search, advertising, and how brands win the digital shelf.