Google’s war on ad fraud just got a whole lot more automated. In 2024, the tech giant swung the ban hammer on a staggering 39.2 million advertiser accounts. That’s more than triple the number from the previous year. Is this a sign of a dirtier web, or just a more efficient Google?
AI: The New Sheriff in Ad-Town
Google credits this dramatic increase to the deployment of Large Language Models (LLMs). Apparently, these AI brains are adept at spotting shady behavior – business impersonation, dodgy payment details, the usual suspects. The claim is they can suspend accounts before they even serve an ad. Optimistic, isn’t it?
Last year Google unleashed over 50 LLM enhancements upon its platforms. Alex Rodriguez, a general manager for Ads Safety at Google, emphasized the crucial role of humans alongside AI in the process. A dedicated team, including members from Ads Safety, Trust and Safety, and DeepMind researchers, were involved in analyzing deepfake ad scams and devising countermeasures.
Deepfakes and Policy Updates: A 90% Drop?
Remember those creepy deepfake ads featuring public figures? Google claims its technical countermeasures and policy updates last year helped slash those reports by 90%. They suspended over 700,000 accounts for this alone. It sounds impressive, but let’s be honest, the bar was pretty low to begin with.
By the Numbers: US Leads the Pack
The US saw the highest number of suspensions, with 39.2 million accounts terminated and 1.8 billion ads removed. India came in second with 2.9 million suspensions and 247.4 million ads nuked. Top violations in India included financial services, trademark misuse, ad network abuse, personalized ads, and gambling and games. Seems like everyone’s trying to make a quick buck, one way or another.
Scam-related violations led to 5 million account suspensions, with nearly half a billion scam ads getting the boot overall.
Elections and Ads: A Minor Distraction?
In a year where half the world voted, Google verified over 8,900 election advertisers and removed 10.7 million election ads. Rodriguez downplayed the impact of election ads on Google’s safety metrics, calling them a “relatively small” part of the overall picture. Make of that what you will.
Fewer Ads, Better Prevention?
While Google blocked 5.1 billion ads and removed 1.3 billion pages in 2024, the numbers were actually lower than in 2023 (5.5 billion ads blocked and 2.1 billion pages removed). Google spin? This decrease shows their prevention efforts are working. Fewer bad ads are being created because they’re catching the bad guys earlier. Or maybe the bad guys are just getting smarter?
Google also restricted 9.1 billion ads in the past year.
The Appeal Process: Because Mistakes Happen (Maybe)
Of course, with this many suspensions, there’s bound to be collateral damage. Legitimate businesses might get caught in the crossfire. Google assures us there’s an appeal process, complete with human review. They’re even trying to be clearer about why accounts are suspended. Apparently, confusing explanations were a problem. Who knew?
Rodriguez mentioned ongoing efforts to improve transparency and messaging around policy rationales, aiming for better clarity for advertisers in 2024 and 2025.
So, is Google actually cleaning up the internet, or just streamlining its ad policing with AI? The answer, as always, is probably somewhere in between. Just try not to get caught in the algorithms’ crosshairs.
Leave a Reply