What you need to know
- Google has revealed a massive attempt by fraudsters last year to submit abusive edits targeting legitimate Google Business Profiles.
- More than 100 million of these edits were blocked, and more than 7 million fake business profiles were removed.
- To police abusive behavior on the platform, the company used a combination of AI models and human operators.
Abusive behavior has become widespread on the web, making the information you find on any online platform subject to scrutiny. This includes reviews on Google Maps, although thankfully the search giant managed to keep information on the service accurate and relevant, aided in no small way by machine learning.
Google revealed that it had blocked more than 100 million abusive edits that targeted Google Business Profiles last year. The company said it used "machine learning models that improved our ability to catch bot activity and unearth suspicious activity patterns."
The Mountain View-based company also removed more than 7 million fake business profiles. A large portion of those takedowns were credited to artificial intelligence, while more than 630,000 were based on user reports.
In addition, Google foiled more than 12 million attempts to create fake profiles and nearly 8 million fraudulent attempts to claim a business profile. More than 1 million accounts were also deactivated for violating Google's policies against online vandalism or fraud.
On top of abusive activities, Google Maps also saw a rise in attempts to use fake reviews to hurt local businesses. Google said it "blocked or removed more than 95 million policy-violating reviews, over 60,000 of which were taken down due to COVID-related instances."
Its ML models also assisted in taking down more than 190 million photos and 5 million videos that were of low quality or violated its content policies.
Google's latest revelation makes it clear that AI is playing an increasingly important role in limiting the spread of shady activities on the internet.
