Amazon Battles Surge of Fake Generative Era Brands; ChatGPT Used to Create Bizarre Product Names
In an ongoing battle against spammers and counterfeiters, Amazon is facing a new challenge: the rise of fake brands in the Generative Era. These brands, primarily created by drop shippers and small Chinese operations, flood the marketplace with knockoff products under peculiar names that often have no apparent connection to the items they sell.
A New York Times article from early 2020 shed light on this phenomenon, featuring brands like Pvendor, RIVMOUNT, FRETREE, and MAJCF. Products adorned with names such as Nertpow, SHSTFD, Joyoldelf, VBIGER, and Bizzliz flooded Amazon with hundreds or even thousands of seemingly positive reviews. These items were not only priced attractively low but also offered free and quick shipping through Amazon Prime.
Gloves are just the tip of the iceberg. Numerous popular search terms return similar results for various products like white socks, iPhone cables, and sleep masks. Brands like JourNow, Formeu, COOVAN, HOVAMP, Binecsies, BSTOEM, MZOO, ZGGCD, and PeNeede dominate these categories. While some products from these brands may prove satisfactory, it remains baffling that the names chosen for them appear anything but serious.
Enter the Generative Era, where creating a fake brand is no longer limited to loosely stringing together a combination of letters. Sellers have resorted to leveraging advanced language models like OpenAI’s ChatGPT to generate brand names effortlessly. Some sellers have even automated the entire process, with no regard for the legitimacy or appropriateness of the names generated by ChatGPT.
Proof of this lazy practice was found by Futurism, who discovered several new brands and product names on Amazon. These brands were named after ChatGPT’s rejection notice, exposing the lack of scrutiny exercised by the sellers. Some of the results were both eyebrow-raising and revealing. For instance, one brand seemed to admit its intention to create a name connected to a trademarked brand, while another brand’s name bizarrely seemed to warn of unethical behavior.
Although Amazon has removed these listings upon discovery, the prevalence of such fake brands remains a challenge. For instance, a beige table from the brand forwillsky named Sorry but I can’t provide the analysis you’re looking for was available for a whopping $1,146.09. Astonishingly, this same brand also presented a Multi-Purpose Product with a strikingly similar name. It appears that trust and safety/content moderation continue to be significant concerns for Amazon, which aims to be a diverse marketplace accommodating all types of sellers.
Ars Technica also stumbled upon additional examples of these fake brands, including a purple product named sorry but I can’t provide the information you’re looking for. Though these listings might disappear quickly, they present yet another hurdle for Amazon in ensuring a safe and trustworthy environment for consumers.
In conclusion, as Amazon fights against an onslaught of fake brands in the Generative Era, sellers are resorting to automated processes and language models like ChatGPT to create bizarre and nonsensical names for their products. This poses an ongoing challenge for the e-commerce giant, as it seeks to maintain the integrity of its marketplace.
