Artificial intelligence applies to every industry in the world. The main question is how to use AI for counterfeit recognition. The world today is an interconnected marketplace, thanks to the Internet. It’s a good thing, but this led to an accelerated increase in available counterfeits. It’s easy to buy a domain, copy a legitimate online store, and market your store on social media. Some fake websites sell counterfeits, and others don’t sell anything at all. Receiving a fake handbag can be tough, but counterfeit drugs kill more than 250,000 children a year. Catching them is like fighting the Lernaean Hydra: for each head chopped off, two grow in its place. E-commerce already uses artificial intelligence for many purposes. For example, visual recommendations offer new benefits to online shoppers. Can artificial intelligence help both retailers and consumers in fighting the counterfeits?
How Counterfeits Affect E-Commerce
The World Customs Organization (WCO) estimates that 7-9 percent of global trading relates to fake products. The figure below shows the revenue generated from various types of illicit trades. You can see that counterfeiting is by far the most profitable. The problem with counterfeits in e-commerce is can’t know until the item arrives. Counterfeit sellers can offer the same images and content as legit stores. Since customers need to wait for delivery, the damage is done as soon as their credit card is charged.
Fake goods sellers don’t care about intellectual property rights. The reason is that few countries have strict laws on the matter. Some countries are yet to come up with the regulation of any online sale. Counterfeiters are not afraid, and their profit keeps on growing.
Advances in the manufacturing process lead to a new problem. Counterfeit recognition keeps getting more difficult. Fakes can hardly be distinguished from legit products. Product quality and durability are questionable, though. Even experts might struggle in recognizing legit items from the fakes.
How Artificial Intelligence Can Help Recognize and Fight Counterfeits
AI fights on both sides. For example, artificial intelligence powers Deepfakes, fake videos that manipulate existing videos. They have famous people saying things they never said. Many people try to hack AI and explore its vulnerabilities.
But, there are good guys as well. Machine learning and deep learning methods can find their use in anti-counterfeiting applications. Artificial intelligence companies use computer vision to recognize fakes. Data scientists design machine learning algorithms to detect details. These details separate legit items from counterfeits.
To a human observer, two shoes can look identical, with one being legit and the other fake. But, data science methods beat the human eye in detecting details. If the fabric pattern is off, deep learning algorithms will see it deviates and report it as a fake.
Credits: Fake Shoe Collectors Share Their Experiences - Complex
Artificial intelligence companies work with retailers and famous brands. The development of counterfeit recognition will soon skyrocket. Here are some examples:
Big Data Anti-Counterfeiting Alliance
Despite being infamous for counterfeit goods, Alibaba recognizes the problem. They are working hard to keep counterfeits off their virtual shelves. Alibaba announced the launch of the Big Data Anti-Counterfeiting Alliance in January 2017. The alliance includes some big names like Samsung, Louis Vuitton, Swarovski, and Huawei.
The alliance uses big data and machine learning to combat counterfeits. “The most powerful weapon against counterfeiting today is data and analytics, and the only way we can win this war is to unite,” said Jessie Zheng, Alibaba Group Chief Platform Governance Officer.
Amazon’s Brand Registry Program
Amazon’s machine learning algorithms helped make Jeff Bezos one of the wealthiest people in the world. But, many brands have criticized Amazon for not preventing counterfeits. Amazon then decided to launch an anti-counterfeiting program.
Brands with registered trademarks can apply to the Brand Registry Program. It helps sellers identify and report infringements of intellectual property rights. The latest updates use machine learning algorithms to detect violators. Before, legit traders had to manually report counterfeits on Amazon.
Entrupy is an artificial intelligence startup for luxury goods authentication. They offer a scanner and an application for their clients. Its mode of operation is very simple. Users scan the handbag, and machine learning algorithms determine if it’s legit. It works in real-time and does wonders for retailers.
They claim their service offers 98% accuracy. Notable retailers use this technology to authenticate goods from luxury brands. Entrupy is a result of computer vision paired with advanced data science.
Why AI Won’t Be Able to Solve The Counterfeit Problem Completely
Despite its best efforts, even Amazon admits they can’t wipe out fake items from their online store. Their 2018 annual report mentions their powerlessness in fighting counterfeits: “We also may be unable to prevent sellers in our stores or through other stores from selling unlawful, counterfeit, pirated, or stolen goods, selling goods in an unlawful or unethical manner, violating the proprietary rights of others, or otherwise violating our policies.”
The problem with counterfeit items is a set of smaller problems. First, counterfeiting can be divided into three categories:
Fake products: A fake seller is trying to sell you something they claim it’s original. For example, many websites are selling new iPhones for extra low prices. Of course, the secret behind this bargain is that the iPhones are fake.
White labeling: The term refers to removing the logo and branding from products. By definition, it’s not illegal as long as the original brand authorizes it. If that’s not the case, retailers take over the product and use it as their own.
Image theft: Online retailers sometimes use images from legit catalogs for their purposes. It can trick shoppers into perceiving the item as better than it is.
Resolving each problem requires a different deep learning algorithm or new data scientists. Artificial intelligence is powerful, and it relies on big data analytics. Machine learning algorithms can use computer vision to detect fakes. Artificial intelligence helps retailers discover why behind the buy, but it can work for the other side.
Despite their reputation, brands mostly use factories with cheap labor in Asian countries. Counterfeiters are getting rich, and they look to upgrade their manufacturing process. Today, they use artificial intelligence to create better copies of original products. In the beginning, fakers damaged both the original brand and angry customers. Counterfeiters can start producing items that will confuse counterfeit detection experts. If that’s the case, people might like them more due to their lower prices. In the end, only the big, multi-billion brands are the victims.
Counterfeiters use whatever means necessary to make customers buy their fake products. The counterfeit market is significant and damaging. Data scientists have been summoned to help combat this issue with artificial intelligence. Notable brands have already expressed their concerns to online retailers. The rise in fake merchandise hurts their profits. Their customers receive low-quality items. Artificial intelligence is a crucial tool for counterfeit recognition for many reasons. Data science algorithms can learn and deal with new types of fakes. They need to keep up with counterfeiters and their methods.
We are still far from real-time counterfeit detectors. They would flag counterfeits as soon as they appear online. Artificial intelligence is not exclusive to brands and legit retailers. It is a weapon used by both sides, to create counterfeits, and to recognize them!