New research from Columbia Law School's Center for Law and the Economy reveals a troubling gap between what AI shopping assistants can do and what retailers actually allow them to disclose. The study tested Amazon's Alexa for Shopping and Walmart's Sparky chatbot and found both tools are capable of identifying misleading “Made in USA” labels on product listings. Yet neither platform uses those capabilities to flag or remove the questionable products, even when the AI recognizes conflicting information on the same product page.
How the Researchers Caught the Fraud
According to the study, researchers at Columbia Law School's Center for Law and the Economy directly tested Amazon's Alexa for Shopping and Walmart's Sparky chatbot. 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. This capability suggests the underlying product data available to these AI systems is sufficient to detect at least some false or misleading origin claims.
The study also found that this kind of fraud is fairly common across both platforms. When researchers asked the bots why their companies weren't acting on the misleading information, neither pointed to any technical limitation. Amazon's Alexa admitted the harm to American brands is real, but explained the issue is easier to ignore until it costs Amazon something directly. That admission highlights a conflict between consumer protection and marketplace profitability.
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. This asymmetry suggests the platforms are making deliberate choices about what information to surface, and consumer safety may not be the top priority.
Why This Matters Beyond Misleading Shoppers
FTC rules require anything labeled “Made in USA” to be all or virtually all made domestically. 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.
The FTC's “Made in USA” standard has been a cornerstone of American consumer protection for decades. Under the rule, a product labeled with the claim must be assembled or processed entirely in the United States, with all significant parts and processing originating in the U.S. Exceptions are minimal and strictly interpreted. When retailers allow third-party sellers to post “Made in USA” labels that conflict with country-of-origin data elsewhere on the same page, they undermine both consumer trust and the legal framework designed to protect it.
The study's findings raise important questions about the role of AI in e-commerce. As AI shopping assistants become more sophisticated, they have the potential to empower consumers with accurate, transparent product information. But this potential is only realized if retailers choose to configure their AI tools to flag inconsistencies. The researchers found the opposite: instead of using AI to reduce deception, some platforms appear to be using it to deflect inquiries about sensitive topics like domestic manufacturing claims.
AI Assistants as Gatekeepers of Information
Amazon's Alexa for Shopping and Walmart's Sparky represent a growing trend of retailers embedding AI chatbots into their apps and websites. These tools are designed to answer product questions, make recommendations, and help shoppers navigate often overwhelming product catalogs. In doing so, they have become gatekeepers of information. If a chatbot refuses to answer a question about a product's country of origin, or redirects the conversation, that decision shapes what consumers can learn before making a purchase.
The study suggests that this gatekeeping function is not neutral. While Amazon's assistant is willing to discuss products made in China, it blocks similar questions about “Made in USA” claims. This selective transparency may reflect underlying data limitations, but the researchers point out that the same AI systems are capable of detecting the inconsistency when asked directly. The gap between capability and action is what makes the findings so alarming.
For consumers, the result is a shopping environment where a label can be prominently displayed on a product page but the AI assistant is programmed not to investigate or expose contradictions. This is particularly concerning for shoppers who rely on AI tools to make quick purchasing decisions, especially during busy shopping seasons or when comparing similar products across multiple sellers.
The Business Case for Silence
Walmart's Sparky reportedly told researchers that the silence on “Made in USA” claims reflects a business calculation rather than a legal one. That statement, while candid, reveals how marketplace incentives can conflict with consumer protection. Third-party sellers are a major source of revenue for both Amazon and Walmart. Flagging or removing large numbers of listings could reduce selection, frustrate sellers, and require additional enforcement staff. Retailers may be reluctant to invest in such measures unless forced by regulators or faced with legal liability.
Amazon's Alexa reportedly acknowledged that harm to American brands is real, but added that the problem is easier to ignore until it costs Amazon directly. This suggests a reactive posture, where the platform might only act if the issue leads to litigation, regulatory fines, or significant negative publicity. The FTC's request last year to crack down on false claims indicates that regulators are aware of the problem, but this study suggests the retailers' response has been insufficient.
Lina Khan and the Battle Against Marketplace Deception
Lina Khan, who founded the Columbia center after her tenure at the FTC, has been a prominent critic of Amazon's marketplace practices. Her decision to sue the company over alleged retail monopoly practices continues to shape the regulatory landscape. This new study adds another layer to that scrutiny by demonstrating that Amazon and Walmart are capable of detecting fraudulent “Made in USA” claims, but appear unwilling to act on that capability.
Khan's argument is that regulators must adapt to the rapid advancement of AI in e-commerce. If AI assistants are going to play a central role in how consumers shop, they should be held to the same standards as human salespeople. A human salesperson who knowingly allowed a false “Made in USA” label to go uncorrected would likely face legal consequences. The study suggests that AI assistants should not be held to a lower standard simply because they are software.
The FTC has already signaled its interest in AI-related consumer protection issues. As AI shopping assistants become more powerful, they may also become the subject of new guidelines or enforcement actions. The commission could argue that a retailer's AI chatbot is an extension of its marketing and customer service operations, and therefore liable for deceptive practices if it fails to correct false claims.
What Amazon and Walmart Say
Amazon responded to the study by saying origin information appears on product pages when available, and that it is working to make that data easier for consumers to find. This response focuses on the presentation of existing information rather than the proactive flagging of inconsistencies. It suggests that Amazon's primary strategy is to improve the visibility of origin details, not to actively audit listings using AI.
Walmart, meanwhile, did not offer a comment on the study. The lack of response is notable given the specificity of the findings. If Walmart's Sparky chatbot is indeed capable of detecting false “Made in USA” claims, the company may prefer to avoid public discussion of why it chooses not to act. Silence in this context may be a deliberate public relations strategy.
Khan says the responsibility now falls on regulators to make sure the public doesn't lose out as AI keeps advancing. Her statement underscores the need for proactive policy measures that anticipate AI's role in consumer protection, rather than reacting after harm has occurred. The study itself is a call to action for regulators to examine not just what AI systems can do, but how they are actually configured and deployed by large retailers.
The findings also invite consumers to be more skeptical of AI shopping assistants, recognizing that these tools are not neutral brokers of information. They are extensions of corporate strategies, and their responses reflect the priorities of the companies that build them. Until regulators clarify standards for AI transparency and consumer disclosure, shoppers may need to verify “Made in USA” claims through independent sources rather than relying on automated chatbots to provide the full picture.
Source:Digital Trends News