Counterfeiting used to be a game of copying popular products. Today, the dynamics of that game have been significantly enhanced to a scale none of us could have imagined. Now, it is a game of replicating the entire customer shopping journey and brand experience.
A counterfeit operator no longer needs to manufacture a convincing fake product and wait for a customer to stumble across it. eCommerce and social commerce, combined with Gen AI technology have completely changed the landscape of brand impersonation.
In hours they can create;
- a polished storefront or website,
- generate or copy professional product photography,
- write persuasive product descriptions,
- imitate brand language and designs,
- manufacture social proof,
- run targeted advertising and build entire networks of seemingly independent sellers
The now-famous AI-generated video of Will Smith eating spaghetti is a useful illustration of this evolution. When the original clip appeared in 2023, the technology struggled with something as simple as accurately depicting a person eating pasta. Just three years later this video is almost flawless. The “Will Smith eating spaghetti” comparison became an unofficial benchmark for just how quickly AI-generative video had improved.
For those of us working in brand protection, that progression matters.
Why? Because it’s the same technology that makes it easier to manufacture convincing digital deception, plus it helps us track, uncover and score it.
The fight against counterfeiting is therefore becoming a tech race: AI against AI, automation against automation, and increasingly sophisticated criminal networks against increasingly sophisticated intelligence systems.
The New Counterfeit Economy
The global counterfeit economy has also been transformed by digital commerce.
The barriers that once limited counterfeiters’ access to consumers — physical premises, distribution networks, advertising infrastructure and budget — have been dramatically reduced. A criminal operator can now establish an online presence in hours, source products, or even components to make a product, through global supply chains and reach consumers in markets thousands of miles away.
More importantly, they can replicate the entire shopping journey a legitimate customer takes.
A consumer might see a product recommended on social media, click through to a professionally designed storefront, read convincing reviews, see photographs that look indistinguishable from official brand imagery, complete a secure-looking payment process and receive a parcel that has been shipped through a legitimate logistics network.
Mostly the customer doesn’t necessarily encounter an obvious “fake” feel to the shopping experience. This is what makes modern counterfeiting so challenging.
The counterfeit is no longer just a product in a box. It is the whole digital ecosystem surrounding it.
Generative AI makes that ecosystem cheaper and faster to reproduce. Product descriptions can be generated automatically. Images can be manipulated or created from scratch. Advertising copy can be localised for different markets. Customer-service conversations can be automated. Websites can be populated at scale.
However, the same principle works in reverse.
AI can analyse abnormal volumes of online activity, score potential threats and categorise counterfeit risks according to a brand’s own priorities and risk appetite. Instead of asking a human investigator to manually review thousands of listings, technology can determine which signals deserve attention first, score and categorise them. That changes the economics and scale of detection and enforcement.
Why Counterfeit Networks Thrive in Today’s Digital Era
Counterfeiters thrive where possible friction is low. Global supply chains, drop shipping infrastructure, print on demand and online marketplaces have made it remarkably easy for a new seller to enter a market without building a traditional retail operation.
Social media has added another layer.
Influencers, viral trends and algorithmic recommendations can turn an innovative or new product into a global phenomenon almost overnight. This creates enormous legitimate demand — but it also creates a ready-made market for counterfeiters to tap into.
TikTok, for example, has demonstrated how quickly shopping and entertainment can merge. Its social-commerce environment has also had to contend with counterfeit products and so-called “dupe” culture, prompting greater investment in intellectual-property enforcement.
For more information on TikToks dupe culture and its impact on brand protection read this article.
Overall, the result is an ecosystem in which counterfeit crime increasingly looks less like traditional street trading and more like a technology-enabled business operation.
"increased sophistication and diversification of counterfeiters are utilising digital manipulation, AI-generated visuals, and algorithmic advertising to blur the line between authentic and fake. The business models adopted by counterfeiters are progressively looking like a technology-enabled business"
The Technologies Making a Difference
Since counterfeit networks are becoming more sophisticated, detection technology has to become more sophisticated too.
1. Artificial intelligence: turning risk into a score
AI allows organisations to move from simple detection towards risk-based prioritisation. Not every suspicious listing presents the same threat. A luxury brand may prioritise high-value infringements. A pharmaceutical company may place safety above commercial value. A consumer electronics brand may care about geographic concentration.
2. Machine learning: learning the behaviour of counterfeiters
Counterfeiters rarely behave randomly. They reuse descriptions or official model imagery to create a sense of authenticity. They create related accounts and often move between the same marketplaces. Machine-learning systems can continually learn from these patterns.
A seller that appears legitimate when viewed in isolation may become highly suspicious when its behaviour is compared with hundreds of other accounts.
This creates a continual learning cycle that can be fed into the tech:
Detect → investigate → confirm → learn → detect better.
Human experts remain essential because confirmed enforcement cases and evidence gathered feeds continuous intelligence back into the system.
3. Image recognition: finding the product that words conceal
Visual recognition is particularly powerful because counterfeiters can change language much more easily than they can change the underlying visual characteristics of a product.
Computer vision can compare product photographs against known genuine products, identify distinctive design elements and detect reused imagery across apparently unrelated listings.
It can also help identify listings where counterfeiters deliberately avoid using obvious brand names.
4. Data analytics and human expertise: connecting the dots
Perhaps the most important technology is not a single algorithm at all. It is the ability to bring disparate evidence together.
This is where technology and brand protection experts become particularly powerful: the tech can identify relationships humans would struggle to see at scale, while the expert can determine whether those relationships actually constitute meaningful evidence and pose a threat to the brand.
Following the Digital Trail
The most effective investigations increasingly follow the counterfeit from the screen to the supply chain.
That means monitoring marketplaces, e-commerce websites and social platforms — but also connecting what happens online with what happens in the physical world.
One interesting example comes from research into large networks of fake online stores. LLM-assisted fake shopping activity surged during peak shopping periods such as Black Friday. Research conducted in 2024 in the run-up to Black Friday found a 110% increase in fake stores between August and October, with tens of thousands using the e-commerce technology platform SHOPYY. The common infrastructure, templates and technical characteristics provided valuable clues that apparently separate storefronts could be connected.
Read more about the increase in domain registrations leading up to Black Friday in our article here.
The goal has become less about simply finding fakes and brand infringements, and more about understanding the networks behind them. It is essential to take this approach to ensure that we can identify the highest-risk activity for the client. This ensures that we can put an enforcement plan in place before consumer attention reaches its peak, and to prioritise:
- disrupting repeat offenders
- protecting consumers from fraud
- protecting search visibility and brand equity
- protecting brand reputation
- providing intelligence to legal teams
- supporting physical enforcement
- identifying emerging threats before they become disruptive
Real-World Example: The Labubu Effect
The explosive popularity of Pop Mart’s Labubu dolls provides a useful illustration of how modern anti-counterfeit technology and enforcement can work together.
Authentic Pop Mart products have built-in mechanisms that help consumers and investigators establish authenticity. Pop Mart’s official verification process uses a holographic label and verification code/QR-based process through its authenticity-verification system.
That creates something valuable from an intelligence perspective: a digital identity attached to a physical product.
At the same time, the explosive popularity of Labubu created an equally extraordinary counterfeit opportunity.
Authorities and brand-protection teams have monitored unauthorised sellers and online accounts, building intelligence around where fake products are being offered and how they move through the market. The scale became impossible to ignore.
In the UK alone, Border Force intercepted almost 259,000 counterfeit toys in 2025, including approximately 236,000 fake Labubu dolls. The UK government subsequently reported that almost 240,000 fake Labubus had been seized at the border.
That intelligence-led approach matters because customs officers cannot physically inspect every parcel entering a country.
They need to know which shipments are most likely to contain the risks.
Risk-profiling databases and intelligence from previous cases can help authorities focus limited resources on higher-risk shipments. In the Labubu case, the scale of the seizures demonstrates what happens when a rapidly growing counterfeit category becomes visible to enforcement agencies.
And the story did something else that technology alone cannot achieve, it became headline news. Once the scale of the counterfeit problem entered the public conversation, consumer awareness increased. The enforcement operation became a warning to consumers as well as a disruption to criminals.
The Challenges and Limitations
Counterfeiters adapt as soon as detection methods become effective. As soon as detection has identified a signal as a well known trait, criminals adapt to it. When a marketplace closes an account, another appears. When a domain is blocked, a new domain is registered.
That is why machine learning cannot be treated as a finished product.
The strongest tech systems have a continual improvement loop involving investigators and brand experts feeding into it. When an expert identifies a new counterfeit tactic, that insight needs to become company-wide intelligence that the system can use in future cases.
The Future of Anti-Counterfeit Technology
The next stage will be less about individual technologies and more about how seamlessly they can work together, with constant updates from experts which is essential to give the intelligence meaning for the brands being protected.
AI will become increasingly reliable at detecting anomalies and prioritising investigations. Image recognition will become more precise. Seller behaviour will become a richer source of intelligence. Network analysis will move closer to real time.
We believe that these three developments could be particularly transformative for brand protection.
1. AI-enabled seller authentication
Today, a seller can potentially establish multiple identities across different platforms.
Tomorrow, marketplaces could use AI-assisted authentication and risk intelligence to establish whether a seller is genuinely independent — or simply another storefront belonging to an already identified network.
That could make the traditional relocation cycle far harder to implement. Breaking that cycle requires platforms to recognise the operator behind the storefront, not just the storefront itself.
2. Smarter social-media enforcement
Counterfeiters are increasingly capable of fragmenting their presence across multiple accounts, pages and storefronts.
Future algorithms should be able to identify those relationships earlier, using behavioural, visual, linguistic and technical signals. The goal should not simply be to remove individual posts, listings or accounts. It should be to collate signals that identify the network creating them.
3. Real-time cross-border intelligence
The ultimate ambition is a connected intelligence ecosystem in which relevant information can move rapidly between brands, marketplaces, technology providers, logistics companies and authorised enforcement agencies.
If a counterfeit network is identified in one country, the intelligence should not remain trapped there.
A seller, domain, product or shipment pattern identified today could become a risk signal for another organisation or location tomorrow.
The technology to make much of this possible already exists. The challenge is connecting it.
- Governments and authorities need actionable intelligence.
- Brands need visibility into threats.
- Brand-protection companies need to connect the evidence at scale and speed.
- Logistics companies need to understand the potential risks moving through their networks.
- Technology providers need to build systems that are explainable, adaptable and accountable.
The counterfeiters are already operating as networks. To stay ahead of them, we need to do the same.
The technology is changing the parameters of the fight against fakes. The key is to use it to not merely find more fakes, but to understand who is behind them, how the network operates, where it is moving next — and how to disrupt it before the counterfeit reaches the consumer.
That is the real promise of technology in the fight against counterfeiting.
Not simply finding the fake. Finding the network. Uncovering their strategy.
Want to see how SnapDragon’s AI-Tech can protect your brand?
If you would like to explore how SnapDragon can accelerate your online brand protection, safeguarding your brand’s Intellectual Property, please do get in touch to schedule a demo.
We would love to hear from you, show you what’s possible, and get you protected.
Get in touch today.
