Google said on Wednesday that it suspended 39.2 million advertising accounts on its platform in 2024 – which is more than three times more than last year – in its latest crackdown on advertising fraud.
Taking advantage of large language models (LLM) and using indicators such as business imitation and illicit payment details, Search Dev said he could suspend the “vast majority” of advertising accounts before the advertisement service.
Last year, Google increased more than 50 LLMs to enhance its safety implementation method in all its platforms.
“Although these AI models are very important to us, very important and they have presented a series of impressive improvements, we still have humans in the whole process,” Alex Rodrig, general manager of ADS Safety in Google, said in a virtual media round table.
The executive told reporters that a team of more than 100 experts gathered in Google, including members of the ADS Safety Team, the Trust and Safety Division, and the Deep Mind researchers. He analyzed the Deep Fake Advertisement scams that add public figures and developed people.
The company introduced technical response and up to 30 advertising and publisher policy updates last year. The company claims that these measures helped suspend more than 700,000 criminal accounts, resulting in the reduction of deep -fuck advertising reports by 90 %.
In the United States alone, Google said it suspended 39.2 million advertising accounts and reduced 1.8 billion ads last year, with key violations linked to the misuse of the advertisement network, misuse of commercial mark, health care claims, personal advertising and false statements.
Google said India, which is the most populous country in the world and after China, has seen a suspension of 2.9 million accounts last year in the second largest internet market after China, Google said, adding that the company also removed 247.4 million advertisements in India, including financial services, trademarks, misuse of advertising, misuse of advertising, advertising and advertising.
Of all the suspension of the advertising account, Google said it had suspended 5 million accounts for scam violations.
Overall, the company removed about half a billion HALF HALF advertising related to scams.
Google also confirmed more than 8,900 new election advertisements in 2024, which saw that half of the world’s population goes into elections, and removed 10.7 million election ads. However, Rodrig noted that the volume of electoral advertising was relatively small than Google’s overall advertising numbers and would not significantly affect its safety measurement this year.
Overall, Google said it had stopped 5.1 billion ads last year and removed 1.3 billion pages. Compared to, that Clinked In 2023, more than 5.5 billion ads and 2.1 billion publisher pages took action.
Google told Tech Crunch that the decrease number indicated its improvement in efforts to prevent it. The company said that by improving the suspension of initial detection and malicious accounts, less harmful ads are produced or reached the platform.
He said the company also banned 9.1 billion ads last year.
The important thing is, mass suspension sometimes raises concerns about the extent to which the company applies its rules. Google said it offers an appeal process that includes human reviews to ensure that it has taken “proper action”.
“Often, some of our messages were not so clear and transparent about the details, about what was rational, or arguing, and sometimes the ad is a bit more confused. We have a group of our policies related to it, because it has 2022,” and its ability to help it is about 2022. Helping… Rodrig said.