Of all the negative articles written about Amazon.com Inc. through the years, one piece in particular stung the company more than most. Claiming that Amazon’s aggressive pursuit of growth had come at the expense of a good shopping experience for its customers, New York magazine this January criticized what it called “The Junkification of Amazon.”
It’s true that Amazon’s online store experience had deteriorated due to unreliable and untrustworthy third-party sellers, many of which are fly-by-night brands — many based in China — with scant regard for quality, reliability and in some cases safety. At its immense scale, Amazon had developed a tumorous trust problem.
I was reminded of the article this week when learning about a new initiative Amazon is working on to give its sellers the ability to generate fake “lifestyle” images of products using artificial intelligence. A tool, currently in beta, takes a boring (real) image of the seller’s item — such as a toaster, say — and spins up a more interesting shot in seconds. Perhaps the toaster will now be on a kitchen counter, next to some tasty looking croissants. The images can then be used in advertising slots on Amazon’s website.
With better margins than its core retail business, advertising has become a critical profit center for Amazon at a time when investors are pressuring the company to boost earnings. In the last quarter, as reported late Thursday, advertising was the fastest growing segment in Amazon’s entire business. Ads have generated $32 billion in revenue this year.
Crucially, unlike some other digital ad players, Amazon said it was not worried about political volatility affecting overall ad sales, because while it was losing some “top of funnel” ads — such as TV slots on Prime Video — it had no such issue with “further down the funnel” sales of sponsored slots within its store. These AI tools are designed to generate even more advertising revenue by helping those advertisers create more ads more quickly, without going to the effort of actually taking a real picture. “Click-through rates can be 40% higher compared to ads with standard product images,” Amazon said in a press release.
The motivation for Amazon and advertisers is clear. For the Amazon shopper, though, the rolling out of generative AI feels like yet more erosion of trust. The company disputes this perception — with reasoning I’ll get to in a moment — but what this comes down to is the potential for more misrepresentation about the quality of products and the integrity of the seller peddling it to buyers.
Amazon, to state the obvious, is not like a traditional store. You can’t use your own judgment over the quality of the products by looking at and touching them in person. Instead, you must rely on imperfect clues: Does it come from a brand I trust? Have lots of other people bought it? Are the reviews good? Over time, Amazon has allowed all of these signals to become messy: the brands are often complete unknowns, sellers can pay to make their products appear more popular, and the reviews system has long been abused.
Generative AI removes further clues. One way quality items on Amazon are able to stand out against cheap knockoffs is because the wonky photoshopping makes it obvious. A fake lifestyle image gives the impression of a more professional outfit, but a seller that can afford to properly market its product — with real imagery — and has gone to the effort to do so is the one I’d prefer.
Another tool Amazon now offers generates product descriptions with AI, a development that will make it harder to spot a low quality product — poorly written descriptions in broken English was a good clue to a bad item. AI will tidy those up in a jiffy.
Amazon stressed to me that the AI-generated lifestyle images are restricted for use in advertising slots only and not product listing pages. I don’t draw any ethical distinction there, personally. The company added that it has a robust moderation process around advertising, but wouldn’t expand on what specific additional measures it had put in place to make sure the AI tool wasn’t being used to give an unfair impression of the product, however slight. It argued that because the products themselves are real, with only the surrounding scene being AI-generated, customers can be confident they are getting what they paid for (and that there’s a good returns policy if they don’t).
This feels like the start of a rather slippery slope. In the examples shown by Amazon, there is no disclosure that AI is being used. Amazon told me it was still considering how labeling might be applied but it should have been in place on day one as a bare minimum.
Amazon founder Jeff Bezos liked to say the company’s guiding principle was to always do what was best for the customer experience. But features like this make me wonder which “customer” Amazon cares more about today? The growing importance of the ads business to maintaining Amazon’s bottom line means the company is starting to behave like other ad-dependent tech giants, for whom the customer isn’t you or I, but the companies who have marketing budgets. The pro-consumer move would be for Amazon to insist every image in its store was genuine, and to use AI to detect and stamp out any kind of fakery. They certainly shouldn’t be encouraging it because that’s just asking for more junk.
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out our most recent white papers.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
