(Credit: Tayfun Coskun/Anadolu via Getty Images)
Getting your ads in front of Google users can mean big business—for scammers, too. Google suspended over 39 million US ad accounts suspected of fraudulent activities in 2024, the company says in an annual ad safety report published Wednesday.
The number is three times more than in 2023 (12.7 million), and the most common violations included ad network abuse, trademark infringement, improper use of personalized ads, false healthcare and medicine claims, and misrepresentation.
To combat threats posed by ever-evolving bad actors, Google says it added more than 50 enhancements to the large language models (LLM) used for detecting its ad policy violations.
These updated AI models require only a “fraction of the information earlier models needed,” relying on signals like business impersonation and illegitimate payment details for early indications of fraud. They contributed to 97% of Google’s enforcement actions, including the removal of 1.8 billion bad advertisements in the US and 5.1 billion globally.
Last year, Google also bolstered its security against deepfake ads used in scams. A team of 100 experts analyzed common trends and updated the company’s Misrepresentation policy. This led to the suspension of 700,000 offending advertiser accounts and a 90% drop in reports of deepfake scams, Google claims.
In a year that saw major elections across half the globe, Google says it verified more than 8,900 new election advertisers and took down 10.7 million ads from unverified accounts.
Restricting legally or culturally sensitive ads was another area of focus. Google restricted 9.1 billion ads in regions where they might be inappropriate; 96 million of those promoted adult content, while 109 million promoted gambling and games, and 34.2 million promoted alcohol.
A major chunk of Google’s annual revenue stems from its advertising business. The company earned $348.16 billion from ads alone in 2024.