Amazon and Walmart AI assistants detect Made in USA fraud but won't flag it, study reveals

Amazon, Walmart AI detect 'made in USA' fraud but do not flag it, study says

July 30 (Reuters) - Amazon (AMZN.O), opens new tab and Walmart (WMT.O), opens new tab AI shopping assistants can often detect when "Made in USA" product labels are false, but the big retailers are not using that technology to crack down on the listings, according to a new study, opens new tab from a think tank led by former chair of the U.S. Federal Trade Commission Lina Khan. The study resurfaces questions about AI shopping assistants' mixed incentives, as retailers see the technology as a way to boost spend. Amazon's Alexa for Shopping and Walmart's Sparky shopping assistants are capable of detecting ⁠mismatches between explicit "made in USA" claims and contradictory information in product listings, the study's authors wrote. When the AI bots are queried on why the companies are not doing more to combat the false labeling, the AI-driven chatbots have cited business justifications. When researchers asked Walmart's Sparky why it does not flag suspicious "made in USA" claims, it responded that the FTC typically enforces "made in USA" rules against manufacturers, not retailers. "That's a business calculation, not a legal justification," Walmart's chatbot said, according to the study. A spokesperson for Walmart did not immediately comment. "Country-of-origin information, when available, is currently displayed on product detail pages, and we're continually working to improve Alexa for Shopping to make this information even more accessible ⁠for customers," a spokesperson for Amazon said. The study is the first published by Columbia Law School's Center for Law and the Economy, launched after Khan returned to the university after her stint with the Federal Trade Commission. "Even as AI tools continue to grow in sophistication and capability, business incentives will shape how these advancements get deployed. Policymakers and enforcers have a vital role ⁠to play to ensure the public doesn't get the short end of the stick," Khan said. FTC rules require products advertised as "made in USA" to be "all or virtually all" made in the United States. The FTC last year urged Walmart and Amazon to crack ⁠down on third-party sellers' "made in USA" claims, citing company policies requiring sellers to provide accurate information. When asked about false "made in USA" claims persisting on the platform, Amazon's Alexa replied, according to the authors, "the harm to U.S.-made brands is ⁠real and documented, but until that harm creates a financial, regulatory, or reputational cost for Amazon specifically, it remains easier to do nothing." Khan sued Amazon when she was at the FTC, accusing it of holding illegal online retail monopolies. The case is ongoing. Reporting by Jody Godoy in New York; Editing by Stephen Coates Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Artificial Intelligence * ESG Regulation Jody Godoy Thomson Reuters Jody Godoy reports on tech policy and antitrust enforcement, including how regulators are responding to the rise of AI. Reach her at [email protected]

Shopping on Amazon or Walmart? Why you shouldn't rely on their AI assistants just yet

Researchers say Amazon and Walmart's AI shopping assistants hide American-made products AI is rapidly becoming the new storefront. Instead of scrolling through pages of search results, more shoppers are asking AI assistants to recommend products, compare prices and narrow down options. But a new report argues that those same AI systems may also be quietly shaping what consumers never see. Researchers at Columbia Law School's Center for Law and the Economy have published a report accusing Amazon and Walmart of using their AI shopping assistants in ways that make American-made products harder to find, despite the systems appearing capable of identifying them. The report, Made in America, Hidden by AI, examined how Amazon's and Walmart's AI shopping tools responded to questions about products made in the United States. According to the researchers, Amazon's chatbot often declined to answer straightforward questions about American-made products while readily answering equivalent questions about products manufactured in China. After rephrasing the same request, however, the chatbot returned American-made products using information that had been available all along. "We tested these systems the way an engineer would, and the refusals fell apart under basic testing," Erie Meyer, senior fellow at the Center for Law and the Economy and one of the report's authors, said. "The data was there, but the design was built to hide it." AI shopping assistants are becoming gatekeepers The report arrives as retailers increasingly push shoppers toward AI-powered recommendations instead of traditional keyword searches. Whether it's suggesting products, comparing alternatives or summarizing reviews, these assistants are becoming the first stop for many online purchases. According to the report, neither Amazon nor Walmart currently offers shoppers a dedicated "Made in USA" filter, despite providing hundreds of other filtering options across many product categories. The researchers argue that if AI assistants become the primary interface for online shopping, the way they choose what information to surface, or withhold, could influence purchasing decisions on a massive scale. I tried it by asking for a few common items like furniture and fishing poles, Alexa+ said it did not have access to the information. However, when I asked for American flags made in the U.S.A., it supplied a product that matched my query. Lina Khan, director of the Center for Law and the Economy and former chair of the Federal Trade Commission, said the findings demonstrate that increasingly sophisticated AI systems are still shaped by the incentives behind them. "Even as AI tools continue to grow in sophistication and capability, business incentives will shape how these advancements get deployed," Khan said. "Policymakers and enforcers have a vital role to play to ensure the public doesn't get the short end of the stick." The systems appear capable of spotting suspicious claims One of the report's more interesting findings is that both companies' AI assistants appeared able to identify potentially misleading "Made in USA" claims. According to the researchers, Amazon's chatbot identified listings where promotional material claimed products were American-made despite evidence suggesting they were imported. Walmart's chatbot reportedly ranked "Made in USA" claims by how suspicious they appeared based on factors including seller information and manufacturing footprints. The authors argue this demonstrates the underlying AI systems have the technical capability to help identify questionable origin claims. However, the report does not establish that Amazon or Walmart internally use these capabilities for marketplace enforcement. Amazon responds In a statement to Reuters, Amazon disputed the broader implication that it hides country-of-origin information from shoppers. "Country-of-origin information, when available, is currently displayed on product detail pages, and we're continually working to improve Alexa for Shopping to make this information even more accessible for customers," an Amazon spokesperson said. Walmart did not immediately respond to Reuters' request for comment. Bottom line The report highlights a question that is likely to become more important as AI shopping assistants continue to replace traditional search. If consumers rely on AI to decide what products to buy, the design choices behind those systems become far more consequential. Decisions about which products are surfaced first, which attributes are emphasized and which questions receive complete answers could shape purchasing behavior in ways that are largely invisible to shoppers. Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok. Finally, you can visit our dedicated Tom's Guide Savings Squad hub for expert help on getting the best products for less.

Amazon and Walmart AI tools detect fake 'Made in USA' labels but don't flag them

A Columbia Law School study finds both retailers have the technical capability to detect false country-of-origin claims but have chosen not to act Amazon $AMZN and Walmart $WMT's AI shopping assistants are capable of detecting false "Made in USA" product labels on their platforms but are not using that capability to flag or remove the listings, according to a study published Thursday by Columbia Law School's Center for Law and the Economy. The researchers focused on Amazon's Alexa for Shopping and Walmart's Sparky chatbot, concluding that both are able to detect when a 'Made in USA' claim in a product title or listing conflicts with country-of-origin details found elsewhere in that same listing. Researchers also found that "Made in USA" fraud appears to be common on both platforms. Asked to explain why their parent companies had not acted on the false labeling, both chatbots pointed to business rationale rather than any technical barrier, the study found. On the question of why false "Made in USA" claims continue to appear on its platform, Amazon's Alexa told researchers: "the harm to U.S.-made brands is real and documented, but until that harm creates a financial, regulatory, or reputational cost for Amazon specifically, it remains easier to do nothing." Walmart's Sparky, when asked why it does not flag suspicious "Made in USA" claims, said the Federal Trade Commission typically enforces such rules against manufacturers rather than retailers. "That's a business calculation, not a legal justification," the chatbot said, according to the study. The study also found that Amazon blocks answers to questions about "Made in USA" products while permitting equivalent questions about products made in China, and that Amazon's own chatbot described the discrepancy as a choice the company made to protect its overseas seller base. In a statement, an Amazon spokesperson said the company currently shows country-of-origin information on product detail pages where it is available and is continuing efforts to surface that data more prominently for shoppers. A Walmart spokesperson did not respond to a request for comment, according to Reuters. Under FTC regulations, a product may carry a "Made in USA" label only if it is "all or virtually all" manufactured domestically. Last year the agency called on both retailers to take action against inaccurate country-of-origin claims by third-party sellers, pointing to the companies' own marketplace rules that obligate sellers to provide truthful product information. The study is the first published by the Columbia center, which was launched after former FTC Chair Lina Khan returned to the university. "Even as AI tools continue to grow in sophistication and capability, business incentives will shape how these advancements get deployed," Khan said in a statement. "Policymakers and enforcers have a vital role to play to ensure the public doesn't get the short end of the stick."

Amazon and Walmart's AI can spot fake 'Made in USA' labels but won't tell you, says study

The same chatbot that happily discusses "Made in China" claims stays silent on "Made in USA" ones. When you ask an AI shopping assistant to help you find products, you probably expect it to point out misleading listings too. According to a new study, both Amazon and Walmart's AI tools can detect fake "Made in USA" claims. However, the study claims that retailers are not using those capabilities to flag or remove the misleading listings, even when the AI recognizes conflicting information on the same product page. How did the researchers catch this fraud? According to Reuters, researchers at Columbia Law School's Center for Law and the Economy tested Amazon's Alexa for Shopping and Walmart's Sparky chatbot directly. Both assistants proved capable of spotting mismatches between bold "Made in USA" claims and contradictory country-of-origin details buried elsewhere in the same listing. The study also found this kind of fraud is fairly common across both platforms. When researchers asked the bots why their companies weren't acting on it, neither pointed to any technical limitation. Amazon's Alexa admitted the harm to American brands is real, but explained it stays easier to ignore until it costs Amazon something directly. Recommended Videos Walmart's Sparky was more blunt, stating outright that its silence reflects a business calculation rather than a legal one. Researchers also noticed Amazon's assistant readily answers questions about products made in China, while blocking similar questions about "Made in USA" claims entirely. Why does this matter beyond just misleading shoppers? FTC rules require anything labeled "Made in USA" to be all or virtually all made domestically, and the agency already asked both retailers last year to crack down on false claims by third-party sellers. This new study is the first from the Columbia center, founded by former FTC Chair Lina Khan, who separately sued Amazon during her tenure over alleged retail monopoly practices, a case still ongoing today. Amazon responded by saying origin information appears on product pages when available, and that it's working to make that data easier to find. Meanwhile, Walmart didn't offer a comment. Khan says the responsibility now falls on regulators to make sure the public doesn't lose out as AI keeps advancing.

Amazon, Walmart AI detect 'made in USA' fraud but do not flag it, study says

Retailer AI assistants can identify false "Made in USA" claims on products. However, these technologies are not actively used to police such listings. A recent study highlights this disconnect, questioning corporate incentives for AI deployment. The FTC previously urged Amazon and Walmart to address these misleading product labels. Business justifications, not legal ones, are cited by the AI chatbots for inaction. Amazon and Walmart AI shopping assistants can often detect when "Made in USA" product labels are false, but the big retailers are not using that technology to crack down on the listings, according to a new study from a ⁠think ⁠tank led by former chair of the U.S. Federal Trade Commission Lina Khan. The study resurfaces questions about AI shopping assistants' mixed incentives, as retailers see the technology as a way to boost spend. Amazon's Alexa for Shopping and Walmart's Sparky shopping assistants are capable of detecting mismatches between explicit "made in USA" claims and contradictory information in product listings, the study's authors wrote. When the AI bots are queried on why the companies are not doing more to combat the false labeling, the AI-driven chatbots have cited business justifications. ⁠When researchers asked Walmart's Sparky why it does not flag suspicious "made in USA" claims, it responded that the FTC typically enforces "made in USA" rules against manufacturers, not retailers. "That's a business calculation, ⁠not a legal justification," Walmart's chatbot said, according to the study. A spokesperson for Walmart did not immediately comment. "Country-of-origin information, when available, is currently displayed on product detail pages, and we're continually working to improve Alexa for Shopping to make this information even more accessible for customers," a spokesperson for Amazon said. The study is the first published by Columbia Law School's Center for Law and the Economy, launched after Khan returned to the university after her stint with the Federal Trade Commission. "Even as AI tools continue to grow in sophistication and capability, business incentives will shape how these advancements get deployed. Policymakers and enforcers have a vital role to play to ensure the public doesn't get the short end of the stick," Khan said. FTC rules require products advertised as "made ⁠in USA" to be "all or virtually all" made in the United States. The FTC last year urged Walmart and Amazon to crack down on third-party sellers' "made in USA" claims, citing company policies requiring sellers to provide accurate information. When asked about false "made in USA" claims persisting on the platform, Amazon's Alexa replied, according to the authors, "the harm to US-made brands is real and documented, but until that harm creates a financial, regulatory, or reputational cost for Amazon specifically, it remains easier to do nothing." Khan sued Amazon when she was at the FTC, accusing it of holding illegal online retail monopolies. The case is ongoing.