From Whack-a-Mole to Intelligence: How AI can Rewrite Anti-Counterfeiting

Written by Paolo Beconcini, IP Expert and collaborator of the China IP SME Helpdesk.

Over the past decades, anti-counterfeiting has often been seen as a resource drain with no clear return. For many, it feels like an endless game of whack-a-mole. There are several reasons behind this frustration.

China, for instance, remains a complex and highly politicized legal environment that is not designed to address counterfeiting efficiently, a topic I have explored in previous articles. At the same time, the resources brands allocate to protection efforts are often limited relative to the scale of the problem. Policing vast digital and physical marketplaces is not something brands can—or should—carry alone; it is fundamentally a government responsibility, and even public authorities struggle given the magnitude of the issue. The result is predictable: a high volume of small-scale enforcement actions, limited deterrence, and a problem that continues to grow year after year. Many brands appear resigned, doing the institutional minimum while accepting the structural limits of what they can achieve.

There is, however, a meaningful shift underway. Artificial intelligence is not a cure-all—it will not replace human investigators or eliminate counterfeiting overnight. But it is already reshaping how anti-counterfeiting efforts are conducted and, more importantly, expanding what is possible when used effectively.

Below are four areas where AI is beginning to make a tangible difference.

Counterfeiters will find it increasingly difficult to hide their digital footprints.

Even an unregistered operation in the remotest parts of central China cannot remain invisible indefinitely. The moment a manufacturer fulfills an order, when counterfeit goods are produced, listed online, sold, and shipped, digital traces are created. AI can collect and analyze these traces, mapping the earliest points of infringement within a supply chain. By linking fragmented data, AI can identify individuals and uncover networks of related entities connected to the original source. It can distinguish the nature of each entity’s business, cross-reference prior cases in similar regions, and detect geographic patterns that point to manufacturing clusters. Within these clusters, areas of interest begin to emerge, locations where an otherwise hidden factory is likely operating.

While AI may not immediately pinpoint the exact facility, it can narrow the search to a defined zone based on overlapping signals and connections. By visualizing complex networks of activity, it provides human investigators with actionable leads that can be pursued on the ground. With the support of experienced attorneys and investigators, these insights are then translated into enforcement actions: warehouse raids, controlled interactions with traders, or pressure applied to logistics providers through legal measures such as cease-and-desist letters. Gradually, critical information surfaces, until the identity and location of the concealed operation are revealed.

This is online-to-offline enforcement at its most effective.

AI will make sense of what you see online

Online monitoring systems process thousands of listings every day. They identify suspicious offers, review them, and initiate takedowns. However, only limited data, often of uncertain value, is retained. As a result, these systems rarely generate information robust enough to convert online infringements into actionable offline cases.

AI can change this dynamic by significantly increasing the proportion of online data that feeds into real-world investigations and enforcement. It does so by enriching listing data with information drawn from across the global digital ecosystem. Starting with basic identifiers, such as seller names, email addresses, phone numbers, and physical locations, AI can match and correlate these with company records, alternative addresses, shipping routes and destinations, past court or administrative decisions, liens, assets, and other available digital traces. By aggregating and analyzing this data at scale, AI transforms fragmented signals into coherent intelligence. It can present complex relationships in a clear, visual format, revealing networks and patterns that would otherwise remain hidden. This allows investigators to move beyond isolated listings and make informed decisions about which of the many online traders are worth pursuing offline.

AI will increase efficiency, reduce cost and accelerate processes

What I’ve described so far is not new in principle. Investigators have been doing this work for centuries, following leads, tracing connections, matching data, and mapping infringing networks. I still remember how proud I have felt looking at the connection graphs we had built in some of our more complex cases. However, it had taken a lot of time, lots of errors and ultimatel;y a lot of money. The real constraints have always been time, accuracy, and efficiency. A persistent question remains: among the countless online traders identified through takedown programs, which ones are worth pursuing as potential entry points to a larger case?

Imagine starting with a single online trader selected more or less at random from a client’s takedown list. At the outset, you have no way of knowing whether you’ve chosen a meaningful target. You follow the trail, but with limited resources and time, you are likely to hit a dead end, a false address, or perhaps a real one that leads only to an anonymous apartment in a suburban area of a second-tier Chinese city. At that point, you face a difficult choice: invest further in a costly field investigation, or walk away. Either way, the decision relies heavily on experience and instinct. However, the more you try to objectively validate a target before moving offline, the more resources you consume, both in hours spent and money invested.

Now consider a different scenario. What if AI had already performed this filtering process for you, analyzing vast amounts of data, eliminating weak leads, and surfacing a refined set of high-potential targets? Instead of guessing where to allocate resources, you begin with a shortlist of opportunities where the odds are materially higher. That shift, from intuition-driven selection to data-driven prioritization, is where AI fundamentally changes the game.

This will obviously result in more efficency, less cost and less time to transition from online to offline actions.

AI will allow brands to cooperate

Brand cooperation in anti-counterfeiting already exists, but it remains sporadic and often economically inefficient. In many cases, collaboration only happens when professional investigators—whose business depends on identifying actionable targets—bring the same case to multiple brands after uncovering overlapping infringements.

Outside of a few initiatives led by major e-commerce platforms, there is no consistent, day-to-day environment where brands can access shared data and collaborate from the earliest stages of a case. As a result, opportunities for coordination are often missed, and efforts are duplicated.

AI has the potential to change this by creating shared intelligence spaces where multi-brand infringements are identified in real time. Listings involving multiple brands can be flagged, enriched with data, and made accessible to all relevant rights holders. Instead of working in isolation, brands can engage with the same set of targets from the outset, supported by a common pool of AI-aggregated intelligence.

This shift would not only accelerate cooperation but also improve cost efficiency. By aligning early, brands can share investigative and enforcement expenses, allowing them to allocate resources across a greater number of cases—particularly those with a higher likelihood of success.