
Technology War: How AI Detects and Prevents Counterfeit Goods Online
On today's social networking platforms and e-commerce platforms, fake and counterfeit goods (dupes) are just a few clicks away from consumers and are often offered for sale at extremely cheap prices. Glasgow-based sportswear brand Define is a typical victim when their designs were copied and illegally branded with logos shortly after launch. Although counterfeit goods exhibit very poor quality – easily torn and severely discolored after just one wash – the biggest concern for businesses is the damage to reputation when consumers may equate the counterfeit product with the standards of the real brand.
To deal with the scale of thousands of infringing posts scattered across many platforms, Edinburgh technology company Snapdragon has applied artificial intelligence to conduct automatic scans. The AI system continuously collects data by analyzing keywords, product descriptions, images and logos to detect anomalies in offer lists.
However, this chase is gradually turning into an "AI versus AI" battle. Counterfeiters constantly change their tactics to bypass automatic scanning algorithms. They proactively blur the logo, misspell the brand name, or tweak images taken from the official website by changing the product color, hairstyle, and makeup of the model. More sophisticatedly, many actors are now using AI-generated content to promote fake products to bypass the censorship system.
Once a post is removed, fraudulent sellers can quickly return with new accounts on other platforms in just a few hours. Therefore, instead of just tracking individual posts, AI technology is now being upgraded to collect wide-ranging intelligence information, cross-analyzing seller and store names across many different markets to completely destroy the distribution network. In the context of increasingly unpredictable counterfeit goods, brand protection organizations are forced to constantly upgrade technology to maintain an advantage in this race.
Reference source: How AI is being used to spot fake goods online | BBC News - YouTube