"Visual Deception" Continues

Written by / Meng Huiyuan

Edited by / Chen Dengxin

Typeset by / Annalee

As Taobao leads the charge against AI-generated fake images, will the quality of online purchases improve for consumers?

It's undeniable that leveraging platform power to intercept AI fake images at the source is the quickest way to see results.

With AI technology becoming more accessible, this low-cost violation model continues to fuel the unchecked growth of inferior products. By synthesizing images with AI, unscrupulous merchants can swiftly create "exquisite" product visuals or plagiarize original designs in bulk, drastically reducing the cost of image theft and counterfeiting. For instance, some clothing merchants simply alter backgrounds and models using AI to pass off original designs as "identical styles" across the web.

Taobao's "ban" on AI fake images undoubtedly serves as a strong deterrent for these merchants. As of now, Taobao has intercepted nearly 100,000 excessively Photoshopped or AI-exaggerated fake images.

However, the widespread proliferation of AI fake images is a long-standing issue of image distortion in e-commerce, albeit with a different manifestation from previous heavily retouched images. The natural limitations of single-platform governance make it challenging to transform the deteriorating e-commerce content ecosystem.

From Heavily Retouched Images to AI-Generated Creations

"I fell for it last month! The AI image of the 'wool coat' looked super high-quality, but what I received was a polyester blend!"

"Many equate AI-generated images with Photoshopped ones. No matter how much a merchant Photoshops, it's still within the same dimension. When hiring models, it's impractical to have only one heavily retouched photo that's unrecognizable. With multiple photos, consumers can generally gauge fabric, color, etc., unlike AI-generated materials that are too perfect to be real. Isn't this like using concept art to deceive people into ordering? No matter how well the image is generated, if it can't be realized in the real world, what's the point?"

"It's increasingly difficult to distinguish AI images with the naked eye. Merchants who take actual photos usually have multiple angles or camera positions. Early AI images mostly had a single photo, and even when edited into videos, they reused the same model pose repeatedly without showing the face. But now, AI can generate not just flat images, but also 3D or multi-dimensional ones."

Image source: Xiaohongshu

While the issue of AI fake images is prominent, the problem of products not matching their descriptions has persisted in e-commerce for years. Consumers have long suffered from image distortion.

From the once-prevalent "seller's show" to the "magically altered show" fueled by AI fake images, the issue of image-product mismatch seems to be an evolution from Photoshopping to AI, sparking a new wave of negative reviews and returns. In reality, it's a trick commonly used by unscrupulous merchants, with AI-generated images being an extreme form.

According to Taobao's ban on AI fake images, an influx of seemingly exquisite yet severely "mismatched" AI images is flooding various platforms, misleading consumers and tarnishing shopping experiences. AI fake images further reduce the cost for unscrupulous merchants to copy styles and steal images across platforms, infringing on the legitimate interests of original brands and merchants, and deteriorating the industry environment.

Taobao's Ban on AI Fake Images

From a consumer perspective, a typical case is the image recognition function of e-commerce platforms combined with AI fake images and pre-sale models.

Some consumers report that AI fake images in the women's clothing section are now very exaggerated, with a single sample image able to search for multiple similar product links. Despite varying titles, prices, materials, and descriptions, the AI images tend to be nearly identical. Some merchants try to bypass platform plagiarism checks with minor AI adjustments (like changing clothing colors, adding watermarks, etc.), but the product pages remain highly similar.

Behind this lies the cruel reality of unscrupulous merchants using AI for low-cost image editing, squeezing the survival space of compliant enterprises. AI fake images reduce the cost for illegal merchants to plagiarize across platforms, leaving good merchants who invest in original design and hire models to take actual photos of products feeling disheartened, ultimately leading to a situation where inferior products drive out superior ones.

Crucially, for many women's clothing items that adopt a pre-sale system, merchants first use AI to create images to gauge whether budget orders meet expected values. If they do, they contact factories for production; if not, they cancel all orders.

In this scenario, regardless of whether the pre-sale item is converted into a physical product and successfully delivered to the user, it's a decisive blow. "Producing based on AI images will likely result in a finished product that differs from the image effect. Even if the buyer receives it, they won't be satisfied. If the order doesn't meet expectations, even if the money is refunded, it's still equivalent to everyone being deceived into contributing sales data for the merchant, with no tangible result after waiting so long."

The root of these issues lies in consumers being misled by AI fake images. When receiving goods that don't meet their psychological expectations, they feel deceived and scammed, leading to either refunds and negative reviews at best or a loss of trust in the e-commerce platform at worst.

Put bluntly, if AI images were the same as actual product images, no one would care whether they're generated by AI. Merchants and platforms are well aware that buyers only care whether the received product matches the online image.

The Root Lies in the Entire E-commerce Content Ecosystem

"This is also part of Taobao's platform governance. A while ago, we took action against malicious wool collectors and some malicious store clusters. These are all about optimizing the business environment to protect the rights and interests of consumers and merchants."

Regarding the current crackdown on AI fake images, a Taobao representative told Xin Kedu that this move aims to guide merchants to operate compliantly. "We actually hope that everyone recognizes that Taobao is the first to tackle this industry chronic disease and then lead the entire industry back to quality or healthy competition."

This indirectly confirms how rampant AI fake images have become. From the platform's perspective, using cases and data to illustrate the adverse consequences of AI fake images is more intuitive.

Common AI Fake Images in Online Shopping

"This Tmall store stole our images. We complained about image theft using portrait rights, but now we see that they're avoiding it by using AI face swapping. Does the platform allow such operations?" "Now the platform has AI-generated images, and there's even more image theft. They directly steal our images to generate AI images, and now we can't even file complaints with the original images?" "My images were modified using AI by peers to publish products. Why wasn't my complaint approved? The human eye can recognize that the hairstyles and poses are the same, and it was modified based on my images..." Many original merchants report that unscrupulous merchants use AI to steal product images in bulk at low cost and imitate popular products.

"The model images had their backgrounds changed directly using AI and were used, with even lower prices and higher sales than ours," said the head of an original men's wear brand. Other merchants have reported that when complaining about infringement of model images using portrait rights, the other party continues to steal them through AI face swapping, with extremely high costs for rights protection.

Therefore, rectification actions that can cover the entire industry still rely on the platform's power to achieve.

According to the relevant person in charge of Taobao, the platform will clearly identify typical scenarios of AI fake images, such as "material or style mismatch," "effect distortion or exaggeration," "non-existent scenes leading to product distortion," and "obvious distortion in cut-and-paste images." It will launch recognition models and comprehensively govern AI fake images on the platform through measures like "source interception," "inventory clearance," and "identification reminders."

Specifically, distorted images will be detected and intercepted in real-time at the product publication end. Additionally, AI synthesis prompts will be added to product detail pages to protect consumers' right to know. As of now, the platform has intercepted nearly 100,000 excessively Photoshopped or AI-exaggerated fake images.

Taobao and Tmall Update Their Interpretation of the "No Image Distortion" Rule

However, from the relevant public opinion on this action, it's evident that some consumers and merchants have a misunderstanding of the platform's new regulations.

"The spring of photographers has arrived, because the system will identify renderings as AI as well." "AI is just a target. Didn't Photoshop normally retouch images before AI?" "3D fake images should also be banned! Especially for furniture and appliances, what you buy doesn't match the description." "The main focus is on fake images, and AI is just one of the popular methods now"... Such views are not uncommon.

This further reflects that the current AI fake image crisis is a systematic challenge facing the e-commerce content ecosystem, with its harm permeating multiple levels, including consumer rights, merchant competition order, and the healthy development of the industry.

"Even if AI fake images are banned, if the actual product quality is poor, it will still feel inferior when received. This has little to do with how the fake images are presented." As a consumer noted, "Don't some images already state that everything is subject to the actual product? So the main thing is to improve the honesty of stores. Otherwise, no matter how much rectification is done, it's just changing the form without changing the essence."

From this perspective, the limitations of single-platform governance are clear: cross-platform image theft, algorithmic traffic bias, low-priced inferior products, and other issues still require industry collaboration.

As Taobao's initiative to the entire industry, only by establishing unified AI usage norms (like prohibiting the falsification of material certifications) and improving rapid complaint mechanisms for image theft can we fundamentally curb technology abuse and promote the return of e-commerce to the right path of quality competition.

But this will inevitably require the participation of platforms, merchants, regulatory authorities, and consumers to reconstruct a new e-commerce content ecosystem centered on authenticity and originality.