A recent paper by six Google braniacs describes the complex system Google has developed to detect and block adversarial advertisements” for the benefit and safety of its users. Brands will be glad to hear that adversarial ads encompass those selling counterfeit goods as well as those that proliferate phishing and malware attacks among others.
Put simply Google’s system combines both machines and human experts to create what Google claims is a highly reliable system for detecting adversarial ads and then blocks them from being served. The human element is used to uncover constantly evolving tactics used by adversaries that cannot be automatically detected by machines. New hand-crafted rule-based models are then folded into Google’s automated models to keep them current.
Now whether this was prompted by Google’s recent settlement regarding illegal Canadian e-pharmacy ads or whether Google had been working on this for awhile the good news is that Google is taking proactive steps to ensure better quality ads for its users. That’s good news for consumers and brands alike.”