How Artificial Intelligence Detects Fake Products?

Last Updated on December 13, 2023 by SPN Editor

Amazon, the global leading e-commerce giant has been harnessing the power of artificial intelligence (AI) and machine learning to weed out vendors attempting to sell fake products on its platform.

Three years ago, Amazon established an internal unit dedicated to combating counterfeit crimes. This team is composed of ex-federal prosecutors, data scientists, and law enforcement personnel from across the globe, all working together to target sellers of counterfeit goods.

The company is committed to closely monitoring any suspicious online activity to safeguard its customers. Kebharu Smith, a former federal prosecutor who now serves as the director of Amazon’s counterfeit crimes unit, emphasized the critical role of their data scientists and automated tools in scrutinizing data continuously.

Smith revealed that Amazon employs AI tools to sift through more than 8 billion seller listings each day. The use of machine learning and data gathering enables them to scrutinize logos and trademarks, comparing them with those supplied by the brands. This process helps them identify potential fake products.

This is crucial because, in some instances, consumers are unaware that the products they’re purchasing online are counterfeit. A recent study conducted by Michigan State University discovered that almost 70% of respondents admitted to being tricked into buying counterfeit items online at least once in the previous year.

Just last month, U.S. Homeland Security Investigations, in collaboration with the NYPD, reported the confiscation of over $1 billion worth of counterfeit luxury items. This marked the largest seizure of its kind in U.S. history. The counterfeit industry is thriving, considered the world’s largest illegal trade, with counterfeit goods worth $2 trillion being sold to consumers every year, as per the National Crime Prevention Council.

The U.S. Customs and Border Protection cautions that the advent of online shopping has transformed the counterfeit trade into a hotbed for criminals peddling potentially harmful fakes. The agency is embracing new technologies, including AI and machine learning-powered X-ray machines, to target smaller packages. This innovative technology will enable the agency to gather and analyze the characteristics of genuine products, using this data to identify and target counterfeits at a speed and scale that far surpasses human capabilities.

Entrupy’s CEO, Vidyuth Srinivasan, expressed his shock upon discovering that all the counterfeit handbags they sampled contained lead and cadmium paint on their surface. Over a decade ago, Srinivasan established his company to assist businesses in identifying counterfeit products. His team has compiled a database of images from genuine and counterfeit luxury items, which employs machine learning to detect fakes in real-time.

Srinivasan’s enterprise, Entrupy, leverages its expanding database to compare in-store photos of an item with images in its repository. This enables both consumers and retailers to identify counterfeits instantly using a handheld device. The company has carved a niche for itself in the luxury accessory and sneaker markets and performs brand verification for TikTok Shop and Goodwill Industries.

Vidyuth Srinivasan believes that the use of AI technology holds immense promise in the battle against counterfeits. Srinivasan explained, “We amassed a comprehensive dataset of authentic and fake products. We trained computers to discern those subtle differences, approximately between 2,000 and 4,000 distinct features on every single image.”

Ashly Sands, an intellectual property lawyer and the leader of litigation and anti-counterfeiting at Epstein Drangel LLP, underscored the gravity of the problem during a recent trip to Canal Street in Manhattan, a location notorious for its black market.

She reminisced, “When I embarked on this journey, roughly 12 years ago, we frequently conducted raids with our clients.”

Artificial Intelligence: The Frontline Defense Against Fake Products

The boom in online shopping has regrettably led to an increase in the prevalence of fake products. From tech gadgets and home decor to clothing and beauty products, imitation items have infiltrated various sectors. Nonetheless, Artificial Intelligence (AI) is stepping up as a formidable adversary in this battle.

AI-Driven Visual and Textual Analysis

A host of startups and tech behemoths are harnessing the power of AI-driven visual and textual analysis to identify fake products. For example, DataWeave, based in Seattle, and Neurotags, based in India, have engineered distinctive solutions that verify products to ensure that only genuine items penetrate the market.

DataWeave employs deep learning to identify and remove counterfeit goods from online shopping platforms. Their AI-driven solution for detecting fake products is trained on millions of product catalog images. The system reportedly boasts an accuracy rate of over 95% in identifying counterfeit items.

Blockchain and AI: A Potent Duo

Tech firms and retailers are merging AI with blockchain to enhance the detection of counterfeits. IBM, a pioneer in AI software development, has integrated blockchain with AI and imaging to authenticate products. IBM Research has created the Crypto Anchor Verifier, an AI-based counterfeit detector that utilizes blockchain and operates on a smartphone. Users can snap a photo of any product, and the app then compares that image against a database in a blockchain ledger to verify its authenticity.

Joint Ventures

Joint ventures such as the Amazon-led International AntiCounterfeiting Coalition (IACC) and Alibaba’s Big Data Anti-Counterfeiting Alliance have also been established to tackle counterfeit or fake products. These alliances underscore the growing acknowledgment of AI’s role in mitigating the issue of counterfeit goods.

AI is turning out to be a revolutionary force in the struggle against fake products. By tapping into the capabilities of AI, companies can safeguard their brand image, guarantee customer satisfaction, and foster a more ethical and equitable marketplace.

How to Detect Fake Products Using Artificial Intelligence

AI-Powered Image Analysis: Firms like Entrupy have designed AI-enabled cameras capable of identifying fake products. These cameras take pictures of the suspected fake products and feed them to an AI algorithm, which cross-references them against a database of genuine products.

AI and Blockchain: The combined strength of AI and blockchain is being utilized to combat counterfeits. Blockchain offers a secure and transparent method to trace a product’s journey from its source to the consumer, while AI can scrutinize this data to spot irregularities that may suggest counterfeiting.

Authenticating Genuine Customer Reviews: AI can also be employed to authenticate customer reviews. Counterfeit reviews can often deceive consumers into purchasing fake products, so AI algorithms capable of identifying such reviews are essential.

AI Cryptographic Technology: Innovative technology has been introduced that imprints an invisible cryptographic signature onto packaging. This signature can be read by an AI system, facilitating swift and straightforward product authentication.

AI and Machine Learning for Customs and Border Protection: U.S. Customs and Border Protection is embracing new technologies, including AI and machine learning, to focus on smaller packages that may contain counterfeit goods.

AI Battle Against Counterfeit Products

As counterfeit products become increasingly sophisticated, we are fast approaching a point where human experts may find it challenging to distinguish between genuine and fake products.

In response to this growing threat, several companies are now offering advanced solutions to businesses and brands. Companies like Entrupy, Red Points, and Cypheme are leading the charge in the cost-effective, large-scale identification of counterfeit goods. They offer innovative technologies that scrutinize various attributes such as materials, colors, and packaging to identify fake products.

IBM Research has made significant strides in this area with the development of the Crypto Anchor Verifier. This AI-based counterfeit detector, which operates on a smartphone, utilizes blockchain technology. To use it, one simply takes a photo of any product. The app then compares this image against a database stored in a blockchain ledger to verify the product’s authenticity.

This database is populated with images of genuine items supplied by the manufacturers themselves. The underlying principle is that every object, including products, possesses unique optical patterns that can be recognized by AI. These patterns are inherent in the texture, color, and design of the materials used to construct the products.

IBM asserts that this technology has wide-ranging applications and could be used to authenticate anything from diamonds and cash to wine and pharmaceutical drugs.

The advent of AI in detecting counterfeit products marks a significant milestone in the fight against counterfeiting. By analyzing unique attributes of products and leveraging technologies like blockchain, AI has proven to be a powerful tool in distinguishing genuine products from fakes.

This not only protects the integrity of brands but also ensures consumer trust and satisfaction. As AI continues to evolve, its role in safeguarding authenticity will undoubtedly become even more crucial, heralding a new era of transparency and accountability in the marketplace.