This summer, Amazon unveiled an artificial intelligence-based feature it had been testing for months to summarize product reviews into short at-a-glance descriptions for customers. A useful tool for some, but it hasn’t been smooth sailing thus far.

Will the technology’s mistakes end up hurting sales? Some say it may. 

The eCommerce giant’s tool uses generative AI to pick out recurring or common themes in existing reviews and creates a summary of them in a couple of short sentences. A Bloomberg report, which analyzed the AI-generated descriptions of several products, found that, in some cases, it generated an inaccurate description or exaggerated negative feedback. 

In turn, some Amazon sellers believe the potential product mischaracterization could hurt sales. 

“If you have a handful of negative feedbacks, and the artificial intelligence model summarizes it like it’s a consistent theme, that’s not fair,” one seller told Bloomberg. “It’s really hard to capture the monetary impact of this, but sellers are not happy.”

But an Amazon spokesperson told Bloomberg that they’ve heard “very few concerns” about the feature (called “review highlights”). On the flip side, the company’s internal analysis found it has led to increased sales, and helped customers find products.

Amazon will continue to refine the technology based on feedback, according to Bloomberg. 

“We care a lot about accuracy, and we will continually improve the review highlights experience over time,” an Amazon spokesperson told Bloomberg.

Other uses of generative AI in business-to-business sales have the potential to boost the bottom line. A Gartner report predicts it could reduce the amount of time spent on prospecting and customer-meeting prep by over 50% by 2026. Find out where distribution stands among AI’s hype.