Google on Wednesday said she suspended 39.2 million advertisers accounts on its 2024 platform – more than three times the number from a year ago – in its latest hit for advertising.
Utilizing large language models (LLM) and using signals such as business impersonation and illegal payment details, the search giant said it could suspend an “overwhelming majority” of advertising accounts before ever serving an ad.
Last year, Google launched over 50 LLM improvements to improve its security implementation mechanisms on all its platforms.
“While these models of he are very, very important to us and have given a series of impressive improvements, we still have people involved throughout the process,” said Alex Rodriguez, a general manager for the safety of Google advertising, on a round table of virtual media.
The executive told reporters that a team of over 100 experts gathered across Google, including members of the Advertising Security Team, the Division of Trust and Security, and researchers from Deepmind. They analyzed Ad Deepfake’s frauds that include impersonations of public figures and developed countermeasures.
The company introduced technical countermeasures and over 30 publisher policies and policy updates last year. These movements helped suspend over 700,000 offensive advertising accounts, leading to a 90% decrease in Deepfake’s advertising reports, the company claims.
Only in the US, Google said he suspended 39.2 million advertisers and removed 1.8 billion advertising last year, with major violations associated with advertising network abuse, trademark misuse, health care claims, personalized advertising and abuse.
India, the world’s most populous country and the second largest internet market after China in terms of users, had 2.9 million account suspensions last year, Google said, making it the second highest after the US, the company also removed 247.4 million ads in India, with the top five policy violations, advertising, advertising, advertising personalized and gambling and games.
Of all the advertiser’s account suspensions, Google said he suspended 5 million accounts for fraud -related violations.
Overall, the company removed almost half a billion fraud advertising.
Google also verified more than 8,900 new election advertisers in 2024, who saw half the world’s population go to the polls and dropped 10.7 million electoral ads. However, Rodriguez noted that the volume of electoral advertising compared to the total Google advertising numbers were relatively small and would not significantly affect his security measurements this year.
In total, Google said it blocked 5.1 billion ads last year and removed 1.3 billion pages. In comparison, it blocked over 5.5 billion ads and took action against 2.1 billion publishers’ pages in 2023.
Google told Techcrunch that falling numbers showed improvements in its prevention efforts. By improving early detection and suspension of malicious accounts, fewer harmful ads are produced or reach the platform, the company said.
The company also limited 9.1 billion ads last year, she said.
Most importantly, large -scale suspensions sometimes arouse concerns how rightly a company enforces its rules. Google said he offers a complaint process that includes human reviews to ensure that he took “appropriate actions”.
“Often, some of our message were not so clear and transparent about the specifics, what was reasoning, or reasoning, and sometimes that they left the advertiser a little more confused. We ended up updating a gang of our policies, as it was linked to it, a bunch of our transparency skills in terms of the advertisers, and why is a great focus on what and why. Part of 2024 “, in 2024”, in 2024 “said Rodriguez.
