Last week, a big dispute erupted between Amazon and Perplexity over AI shopping tools. Amazon sent a stern legal letter to Perplexity, demanding changes to how its AI shopping agent operates. The core issue is that Perplexity’s Comet AI browser feature allows an AI “agent” to shop on Amazon on behalf of users. For example, if someone wants to buy cat food, the AI can go to Amazon, find the product, and place the order without human intervention.

Amazon isn’t happy about this. They want the AI to clearly identify itself as an AI agent and not pretend to be a real person. Amazon argues this lack of transparency could let the AI bot bypass their controls. Essentially, Amazon wants to be able to block these bots easily, similar to how media companies try to prevent content scraping by AI.

The tone between the companies quickly heated up. Perplexity accused Amazon of “bullying” and claimed the retail giant is anti-innovation. They said Amazon just wants to push ads and sell unnecessary products. But this dispute touches on something bigger. AI shopping agents could reshape how the digital economy works, especially for Amazon.

Amazon makes a lot of its revenue from ads—over $600 billion last year. These ads help products get noticed in search results and recommendations. Amazon also relies heavily on upselling, where they suggest related items based on your previous purchases, encouraging more buying. They push Prime memberships and other perks that keep customers coming back. If AI agents buy things directly without engaging with these ads or recommendations, Amazon loses a big part of its revenue and customer data.

This battle isn’t just about Amazon. Many online stores rely on personalized offers, discounts, and recommendations to attract and retain customers. Some European retailers are excited about AI shopping agents and chatbots bringing new visitors to their stores, similar to how Google search does. But the problem is, when AI agents act as shoppers, they might bypass all the human interactions that generate revenue for retailers.

The future size of this issue depends on how AI companies and platforms handle these concerns. AI providers like Perplexity might need to respect store rules or negotiate revenue sharing. But some, like Perplexity, have a history of skirting the edges—such as reading paywalled articles without paying—so compliance isn’t guaranteed.

What about users? Perplexity claims that AI shopping agents are what users now want, viewing it as a natural step in technology’s evolution. But not everyone agrees. Many people still enjoy searching for deals, comparing options, and building trust with brands. They prefer making their own choices rather than handing over their shopping to an AI bot, especially when it comes to sensitive transactions like entering credit card info.

Still, others might prefer fully automated shopping, where the AI handles everything from browsing to checkout. If that becomes mainstream, it could have major consequences for online retail. Retailers might see fewer visits and less engagement, which could impact their sales and marketing strategies.

Looking ahead, the rise of AI shopping agents presents both opportunities and risks. Retailers and AI companies will need to figure out how to coexist. Will AI bots enhance shopping experiences, or will they undermine traditional online retail models? The ongoing debate between Amazon and Perplexity is just the start of a much bigger shift coming to e-commerce.

Inspired by

• https://www.computerworld.com/article/4091218/ai-agents-are-bad-e-commerce-customers.html

Sources

• gmpg.org
• aboutamazon.com
• aboutamazon.com
• perplexity.ai
• computersweden.se