Those Peddling Counterfeit Products May Have Met Their Match in Artificial Intelligence

The menace of counterfeit products continues to threaten global communities and keep specialized enforcement agencies on their toes. According to a report published by the United States Secret Service for the financial year 2023, as much as US$22 million worth of counterfeit currency was seized in the year 2023. During that year, the US Secret Service made close to 200 arrests for counterfeiting. The United States Department of Treasury estimates that counterfeit bills – worth anywhere between US$70 million and US$200 million – are in circulation at any given time. 

Counterfeiting is a malice that does not only plague currencies or the circulation of monetary bills worldwide. It affects almost every supply chain. Reports suggest that counterfeiting and piracy cost the global economy as much as around US$600 billion each year.

As per a survey conducted by CRISIL, close to 90% of consumers opt for counterfeit products, lured by the fact that they are cheaper and available extensively. The urge to buy these products also stems from the need for social validation and recognition that many of these expensive and luxurious brands bring to people. 

Amidst this environment of having counterfeit products as part of our daily lives, it is obvious that law enforcement agencies would try to stop it. Any threat to the global economic order would be met with severe consequences. However, what seems to be an interesting development at this stage and point in time is that the efforts to dissuade counterfeiting have been embracing cutting-edge, modern-day technology paradigms. One such technology has been Artificial Intelligence. 

Click here to learn all about investing in artificial intelligence. 

Image Mining Techniques and Machine Learning Algorithms to Identify Counterfeit Coins

A group of Canada”s Concordia University researchers have published a unique framework that leverages image mining and machine learning algorithm techniques to identify flaws leading to the singling out of counterfeit coins. The publication came out in the journal’ Expert Systems With Applications’. The researchers were associated with the Centre for Pattern Recognition and Machine Intelligence (CENPARMI). 

While briefing on how the framework identifies the flaws, Ching Suen, a professor in the Department of Computer Science and Software Engineering and the paper’s supervising author, had the following to say:

“Using image technology, we scanned both genuine and counterfeit coins so we can look for anomalies that may be either two- or three-dimensional, such as letters or the face of the person on the coin.”

Mariam Sharifi Rad, a postdoctoral fellow at CENPARMI and the paper’s lead author, stressed that the framework was not only about safeguarding the economy and its resources but also about ‘pushing the boundaries of technology and improving security.’ 

Here, we must remember that the issue of counterfeiting is deeply intertwined with cybersecurity aspects, especially in this digital age, when virtual transactions are ubiquitous. 

The United States Secret Service report that we’ve already cited to point out the volume of counterfeit currency seized says that in 2023, the amount of loss recovered from crimes related to cyber-finance was as high as US$1.1 billion. The number of arrests made for cyber-financial crimes was 1,019. 

Undoubtedly, these numbers validate the claims the researchers are making about their framework being something that will help fight the larger war and push the envelopes of security and protection against counterfeiting as we’ve known them so far. 

But how did the researchers achieve their mission on the ground? How were their intentions made a reality? For that, we would have to delve deeper into the process. 

Leveraging Fuzzy Concepts, Semi-Automatic Image Mining and PSO

The researchers made judicious use of these three tools to detect counterfeit coins. They created six-coin image datasets and curated them as central references for their research. The deployment of the image mining system depended on fuzzy association rules and employed a blob detection method and relationship predicate. The combinatory system could extract implicit information from images that aligned with human perspectives. 

The framework also integrates the technique of particle swarm optimization (PSO), which aids in dynamically redefining threshold parameters, enhancing the efficiency of the fuzzy association, and making the counterfeit coin detection system more robust. The result is a system that is highly adaptable to an array of varying image datasets. 

To summarize the achievements of the framework, Ching Suen said:

“This method can be used to detect all kinds of fake goods, which we see all over the world. It can also be used to detect fake labels on fruits, wines, liquor, and so on. There are many places where this can be applied.”

There is no doubt that this novel framework to detect counterfeit coins—if replicated and scaled up well—would help the global economy significantly. However, another solution that has been successfully dealing with counterfeit currencies since its inception is blockchain-based currencies like Bitcoin. 

The Property of Immutability in Blockchain-Based Currencies

Blockchain-based currencies and their transactions run on the principles of an immutable ledger. It refers to a system of record keeping that can not be altered.

In the blockchain system, each information block that contains facts or transaction details is backed by a corresponding hash value. Based on highly secure cryptographic principles, this hash value comprises alphanumeric strings generated by individual blocks.

Each block has a hash value or digital signature for itself and the previous one, ensuring an unrelenting and retroactive coupling of blocks that prevents interference with the system or the alteration of the data that has already been saved. 

What makes counterfeiting a redundant concept in blockchain-based currencies is that the hash value cannot be reverse-engineered. No user can utilize the output string to reach the input data. 

Overall, in blockchain-based currency systems, it is nearly impossible to alter or delete the block’s data placed in the blockchain. Whenever anyone attempts to make a change, the modification is rejected by the subsequent block since the hash of the league would no longer be valid.

Apart from currencies, blockchain’s mechanism can deter fraud and counterfeiting in many other aspects of everyday production systems and supply chains. The European Commission, for instance, has advocated for the use of blockchain to secure our global supply chains from the negative impacts of potential lawsuits, consumer injuries, loss of sales, and long-term reputational damage.

Blockchain as a Solution to Combat Counterfeiting across Businesses Worldwide

The European Commission has recognized the role that blockchain technology can play in improving traceability and conducting an end-to-end tracking of products and shipments across a range of supply chains, starting from the procurement of raw material up to the phase where the user gets the finished product. 

The implementation of blockchain may happen through smart tags, and businesses may use many sorts of smart tags, including QR Codes, RFID tags, and signatures on metallic or ceramic surfaces. 

Each of the abovementioned tagging mechanisms has its advantages. For instance, QR Code-powered systems allow businesses to transfer information such as the date of manufacture, the company’s website, or customer care number. They also aid in executing payments and seamless shipment tracking. 

When affixed to products, RFID tags help the reader receive signals deployed along the monitored areas. The system can track movement with high efficiency. However, the system is often seen as unrealistic when the scope of deployment is large and the cost becomes significantly higher. 

As an advanced solution, laser marking machines have appeared in the market. These machines help incorporate traceable barcodes and graphics onto metallic or ceramic surfaces. 

Apart from the European Commission’s deliberations on blockchain and its role in securing supply chains, a paper presented at the 7th International Conference on Computer Science and Computational Intelligence 2022 looked into the possibilities of developing an anti-counterfeit system using blockchain technology. 

The paper put forward blockchain technology as the backbone of a system where customers can validate a product’s legitimacy without the need for a corresponding merchant. The authors of the paper highlighted the potential of the Ethereum blockchain in helping to build the proposed model that could trace every item’s creation and transactions, ensuring the credibility of an item’s genuineness. 

The role of blockchain in securing supply chains and innovations around it has made it clear that our supply chains need protection. Having robustly protected supply chains could mean a sure-shot winner as far as protection from counterfeiting is concerned.

Several companies have been investing research and resources in using Artificial Intelligence to secure supply chains. In the following segments, we discuss a couple of such companies and their solutions. 

#1. IBM

IBM claims that its solutions help build disruption-minimized, resilient, sustainable supply chains. The IBM Sterling® Supply Chain Intelligence Suite, for instance, is an  AI-based optimization and automation solution for helping organizations solve supply chain disruptions through traditional transformation. The solution can transform supply chains digitally, improve resilience and agility, and accelerate time-to-value through actionable insights, smarter workflows, and intelligent automation.

The solution, according to the benchmarking numbers published, reduces the time it takes to trace items from store to farm from 7 days to 2.2 seconds. It reduces inventory levels by 18% to save on waste and cost. Additionally, the time needed for critical supply chain disruption management is shortened from days to hours.

IBM claims its solution to be fit for a range of use cases, including food safety, multi-tier supply chain mapping, tracking and tracing, supply chain emissions management, supplier risk and compliance assessment, supplier visibility, and more. 

Many companies use IBM’s solutions. For instance, Antonello Produce uses it to improve produce traceability from seed to store to shelf. Pietro Coricelli leverages the solution to optimize food quality, sustainable sourcing, and supply chain transparency. Sonoco uses it to safeguard its product quality and the efficacy of its life-saving medications. ‘

The IBM team continues to work on the solution to improve it incrementally each day. For instance, it recently leveraged AI to design a control tower solution that connects existing inventory solutions and enterprise resource planning systems. 

For the fiscal year 2023, IBM generated $61.9 billion in revenue, up 3% at constant currency, and $11.2 billion of free cash flow, up $1.9 billion year-over-year. Scaling AI to enhance productivity within IBM expanded the company’s profit margins. 

#2. Oracle

In April last year, Oracle introduced AI and automation capabilities to help customers optimize their supply chain management. It brought in new capabilities across Oracle Fusion Cloud Applications Suite to help customers with accelerated supply chain planning, increased operational efficiency, and improved financial accuracy. The specific updates included new planning, usage-based pricing, and rebate management capabilities.  

According to Jon Chorley, senior vice president of supply chain applications and chief sustainability officer at Oracle:

“With Oracle’s complete suite of integrated applications, organizations can manage supply chain data on the same platform as finance, HR, and customer experience to accelerate processes like quote-to-cash and remove barriers that have traditionally existed between different business functions. This holistic approach creates an environment where AI and automation can flourish to help businesses drive efficiencies and achieve more with less.”

In March 2024, Oracle announced fiscal 2024 Q3 results. Total quarterly revenues were up 7% year-over-year in both USD and constant currency to $13.3 billion. Q3 GAAP operating income was $3.8 billion. Non-GAAP operating income was $5.8 billion, up 12% in both USD and constant currency.

The Use of AI for a Secure Future

Putting an end to the menace of counterfeiting, fraud, and maliciously produced duplicate materials is something that will see significant improvements from the use of AI-based solutions.

What is more enticing is that the use of AI in securing the supply chain would benefit other areas of an operational framework. For instance, it would make warehouse management more efficient, operating costs would come down significantly, the scope of waste generation and unwanted emissions would be reduced, and inventory management would be more accurate. 

Apart from getting genuine goods, consumers will benefit from more timely deliveries, and workers will be safer. However, companies will have to invest more in making AI more user-friendly by allocating more resources to training and simplifying complex operations by integrating diverse elements across the many nodes of a global supply chain. 

They must also regularly monitor these systems and their performance, identifying and fixing glitches as they arise.

Click here to learn what makes 2023 the breakout year for artificial intelligence.