How AI will stop counterfeit goods once and for all

The rise of the internet made a lot of gruelling tasks a lot easier and simpler, one of the largest examples: shopping. E commerce has become part of the daily lives of millions of people due to the simplicity of the process. However this distancing from the real world tasks that make going shopping in real life a nuisance has also allowed the sale of counterfeits goods to be easier and more profitable than they have ever been.

As website building tools become more widespread and easier to use the creation of a legitimate looking website selling a legitimate looking product has never been easier, in fact it can be done by anyone at near 0 cost. Automation techniques in the real world have also improved to the point that actual counterfeit goods are so close to the originals that the ordinary eye couldn’t tell and at some point in the near future the actual creators will be confused as to which good is fake and which is real.

The problem is begging to catch official attention. In Amazons annual report, the company warned investors for the first time ever about the issue of counterfeits on the platform. The report lays out the following:

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

This has lead to many brands asking themselves if the current system on Amazon’s platform is good enough as most firms cant devote the necessary resources to constantly searching for fake versions of their product. Amazon has responded to these concerns in various attempts to deal with the issue. One being its Brand registry system which hasn’t seemed to have the desired effect they intended.

Traditional counterfeit market in Beijing

Traditional counterfeit market in Beijing

AI,however, may hold the solution to this ever-growing problem. In China this approach has already been implemented. In 2017 Alibaba and 20 other large firms , created the “Big Data Anti-Counterfeiting Alliance”. The goal being to use AI to to go through customer reviews and product listing and spot patterns that might indicate a product is counterfeit. The attempt was successful and lead to the closing down of 230,000 IP-infringing stores on Taobao. Amazons own attempt was announced this year, known as “Zero“ it will use a combination of machine learning self-servicing and product serialisation to help tackle this issue.

Although in its early stages it is not hard too see anti counterfeit good systems fully implemented in the near future. The speed at which this will happen also depends on how wiling brands are too help in the endeavour, by either providing data or working on the systems themselves. However much like the recent news on “deepfakes” we may see equivalents in the counterfeit industries where ML systems will design products out of the cheapest materials while remaining as similar as possible to the original product ending in an arms race much like the one between “deepfakes” and “antideepfake” tech.


Finance, Enterprise, AmazonArturo Dezon