Google blocked more ads in 2023 than in the previous year. AI has led to more deepfakes. At the same time, technology also plays a central role in the fight against fraud.
Last year, Google blocked or removed more than 5.5 billion ads that violated the online company’s policies or were intended to be fraudulent. That’s according to a Google ad security report published in Mountain View.
The blocked content includes advertising for dangerous products or sexual content, but also advertisements that are dubious or fraudulent and promise a miracle cure or quick wealth. Last year, more abusive advertisements were removed than in the previous year. In 2022, Google said it removed around 5.2 billion ads.
Google wants to block pages before they are displayed
Google manager Duncan Lennox, vice president of ad security at the search engine operator, said: “Google’s goal is to identify fraudulent ads and block accounts before they reach Google’s platforms , or to remove them as soon as they are discovered.”
The company does not communicate the number of advertising placements per year on its own pages or via Google’s advertising network. Industry experts estimate that Google’s servers deliver an advertisement an average of around 30 billion times every day.
90 percent of the Block decisions with AI
The most important trend in 2023 was the impact of generative artificial intelligence, as known from chatbots like Google Gemini. This new technology has triggered relevant and significant changes in the digital advertising industry – from performance optimizations to image editing. “But there is also no question that the introduction of readily available AI video tools has exacerbated the proliferation of fraudulent ads using so-called deepfakes,” Lennox said.
Deepfakes are photos, videos or audio files that are intentionally altered using artificial intelligence. You see or hear people doing or saying things that they have never actually done or said. The Google manager emphasized that AI helps to simultaneously detect fraudulent content and enforce Google’s policies. AI is significantly involved in 90 percent of blocking decisions.
12.7 million accounts suspended
If advertisers believe the AI has made a mistake, Google staff will review it. In cases where errors are encountered, these findings are used to further improve the systems.
According to Lennox, in 2023 Google not only removed 5.5 billion questionable or even fraudulent ads, but at the same time blocked or removed 12.7 million advertiser accounts, almost doubling compared to the previous year.