Amazon’s Latest Challenge: When Bots Do the Buying

0:00What we noticed more is that we've moved AI off the screen into the physical world. That's Tanguy Katlan, senior partner at McKinsey and director of the McKinsey Global Institute. AI, he says, no longer lives in a computer but is tangible in robotics, engineering, space travel and more. One of the 14 key points identified by McKinsey in its 2026 Technology Trends Outlook. At the break, join McKinsey to learn how these trends are shaping innovation and how leaders should respond.

0:33Welcome to Tech News Briefing. It's Wednesday, November 4th. I'm Julie Chang for The Wall Street Journal. Amazon has been the online shopping destination for decades now. And being America's go-to virtual shop has helped it rake in tens of billions of ad revenue annually. But there's a new threat that could be putting those ad dollars at risk. AI agents. Take Meta's Muse, which has signed up millions of users since its launch in September. But why are these virtual assistants posing such a problem for Amazon? Here to break it all down is WSJ reporter Isabella Simonetti. Isabella, let's start with how big is Amazon's ad business?

1:19Amazon's ad business is one of the company's biggest cash cows, and it brought in almost$70 billion in revenue last year. There are a variety of different ways that Amazon makes money off of advertisements, but a big chunk of it comes from charging companies to play sponsored ads in front of Amazon shoppers. So then how is Muse able to disrupt Amazon's advertising business? So Muse has the potential to disrupt Amazon's advertising business because it essentially bypasses the human buying experience. It finds and buys products, whereas before you might have browsed Amazon.com and AI agents can evaluate product specs and details to show you what to buy rather than, you know, if I'm going to Amazon.com, I'm taking in advertising that might sway my behavior.

2:17So if an AI agent makes a purchase, a brand may have less incentive to pay Amazon for ads that aren't having as much of an impact on what consumers buy. So just to be clear here, my Muse agent, because it's shopping on my behalf, Amazon is not getting human eyeballs looking at the ads on its site. And that's why Amazon wouldn't be getting any of those ad dollars. That's pretty much right. Yes. So we've been talking a lot about AI agents recently, but are people really shopping with their AI agents? Muse has become an instant hit, getting millions of downloads and people setting up agents really early on into its tenure.

3:01But most people are not using agents to buy products for them. I mean, think about your everyday habits that are ingrained in your life. You might go to Amazon or you have certain stores that are go-tos for you. But in AI, when a week can be a year in terms of technological development, companies like Amazon that are a first stop for e-commerce shoppers want to be on the forefront of an agentic shopping technology. J.P. Morgan analysts said in a recent note that Muse has the potential to become the most widely used consumer AI application since ChatGPT. Okay, so what has Amazon done in response to the presence of Muse or other AI agents that can shop on behalf of their humans?

3:52Shortly after Muse was released, Amazon blocked Muse from shopping its site and people noticed very quickly. Amazon put out a statement saying that they were blocking Muse, citing safety concerns. Andy Jassy even brought up the advertising issue on Amazon's earnings call last week, saying that outside AI agents aren't yet eating into Amazon's advertising business. But he was talking up Amazon's own AI shopping feature, Alexa for shopping, saying that that's going to be the main shopping agent customers use to shop Amazon. Have you used an AI agent to shop online? If you're a listener on Spotify, let us know in the comments.

4:40Coming up, what steps can Amazon take in its fight against outside AI agents on its platform? That's after the break.

4:57AI is delivering tremendous advances in areas like scientific discovery and engineering. But as Tanguy Katlan, senior partner in McKinsey and director of the McKinsey Global Institute notes, breakthroughs only happen when leaders set the table effectively. You can only optimize and improve the performance of the system if you improve every single component of the system. To really unlock the full potential of those technologies, you're going to need to start from a blank sheet and redesign the processes and trends. Later, Katlan explains how leaders identify the trends that matter most to their businesses.

5:30Welcome back. We've been talking about how personal AI agents could disrupt ad revenue for Amazon. Isabella, have you tried out Alexa for shopping? I have tried out Alexa for shopping, yes. And what did you think of it? Does it work? It works. It's just different than Muse. So Amazon has a buy for me feature, which for some products allows you to kind of buy them directly. Amazon has introduced some new features for Alexa for shopping recently. and they say that more than 400 million customers have used it in the last 12 months, which is a pretty substantial number. But for me personally, the way I shop on Amazon is already pretty ingrained and set in stone.

6:22I browse, I know what I want, I go search it and I buy it. And I'm not trying to necessarily do really detailed product comparison or have something buy something for me unless I'm looking for something really specific. So then what do you think it'll take for Alexa for Shopping to win back users who may have already adopted agents like Muse? Amazon isn't going anywhere. Most of the analysts and experts who I spoke to have said it's really difficult to disrupt Amazon's business model. And more than that, Amazon's place in people's everyday lives, So many people rely on Amazon for groceries, essentials.

7:03So I don't think that people are necessarily abandoning Amazon in favor of Muse. I think for a lot of people, Muse is sort of an add-on. So if you're trying to buy a bike or create a budget, Muse is sort of a personal assistant for you. And Andy Jassy hinted at this on the Amazon earnings call saying that he anticipates that Amazon will be working with outside agents. So it's possible that Amazon could cut a deal with Muse. So according to the experts you spoke with, how much of an impact could these AI agents really have on Amazon's bottom line? Because you just mentioned Amazon's not going anywhere.

7:45Right. It has a very difficult business model to disrupt. And I think the impact of shopping agents on Amazon's business, it's hard to tell because it's pretty early days. Muse only came out in September and it's been a pretty quick hit and is very popular. But whether or not that can have a substantial impact on Amazon is yet to be seen, especially when barely two months after Muse came out, the CEO of Amazon is talking about potential deals with outside agents. The experts that I spoke to talk about how in the past Amazon has adapted pretty well to big changes like this and made their own plays for the future of whatever the challenge was at the time.

8:30And now that's agentic commerce. But there are some outside spectators like Elon Musk or the Silicon Valley investor, Paul Graham, who had said this is an opportunity to launch an Amazon competitor, which is pretty interesting. What is the challenge Amazon is facing in this new era of AI shopping agents? One of Amazon's core values and principles is customer obsession. So focusing on customers and not competitors. But when you have customers or at least a lot of people who are really into muse and outside shopping agents, how do you deal with that? Especially when e-commerce is a big part of your business.

9:14That's the main challenge for Amazon here. That was WSJ reporter, Isabella Simonetti. And that's it for Tech News Briefing. If you're a listener on Spotify, be sure to leave us a comment. Today's show was produced by me, Julie Chang, with supervising producer Katie Ferguson. We'll be back tomorrow with a new episode. Thanks for listening.

9:43Autonomy, absorption speed, cybersecurity, co-design of chips and models, unlocking the capacity that we need in the grid and energy. That's McKinsey's Tanguy Katlan, discussing five key themes outlined in McKinsey's 2026 technology trends outlook, a comprehensive view of what's shaping today's tech landscape and where we might be tomorrow. Leaders, Katlan says, can examine the 14 trends McKinsey identifies and apply the learnings to their unique business context without chasing headlines. The competitive advantage comes from the intersection of proprietary data, workflows, and the ability to rewire the operating model, not from the trend that gets the most capital or gets the most attention to the press.

10:24It's how you're going to redesign your end-to-end workflows, take advantage of your own proprietary context and data, and the ability to rewire the operating model at upskill.com. That's where actually you win. For more, visit McKinsey.com. This content was created by Custom Content from WSJ, a unit of the Wall Street Journal Advertising Department.