Meta $META ( ▲ 0.32% ) derived roughly 19% of its $18 billion in advertising revenue from China last year from ads tied to scams, illegal gambling, pornography, and other prohibited content, according to a new Reuters investigation based on internal company documents. The report follows an earlier Reuters probe that found about 10% of Meta’s global ad revenue came from similar categories, underscoring how important Chinese advertisers have become to the company’s business.
China plays an outsized role in Meta’s ad engine even though Facebook, Instagram, and WhatsApp are banned inside the country. Internal documents reviewed by Reuters estimate that more than $3 billion of Meta’s China revenue in 2024 came from what employees internally dubbed a “scam tax.” Meta also identified China as the top source of scam ads globally, with victims including US and Canadian investors and Taiwanese consumers buying fake health products.
An Anti-Fraud Crackdown That Did Not Last
Meta attempted to rein in the problem in early 2024 by standing up a dedicated anti fraud team. New tools briefly cut the share of problematic Chinese ads from 19% to about 9% in the second half of the year.
But after CEO Mark Zuckerberg reviewed the effort, the company executed what documents describe as an “Integrity Strategy pivot.” The anti fraud team was disbanded, a freeze on new Chinese ad agencies was lifted, and by mid 2025 the rate of banned ads had climbed back to roughly 16% of China revenue.
How the Machine Keeps Running
The investigation details how the system works. Because Meta has no direct operations in China, it relies on a small group of large reseller agencies that sell ad space to layers of smaller partners. To protect revenue, some top resellers are “whitelisted,” meaning their ads are not automatically removed when flagged and instead go through slower human reviews. That delay allows scam ads to run for days.
Reuters reporters were able to buy ads for a fake get rich quick scheme for under $30 using a lower tier partner, with the ads running without interruption.
Internal dissent was sharp. Former Meta executive Rob Leathern reportedly called the fraud levels “indefensible,” questioning how anyone could view them as acceptable. Documents cited by Reuters show employees explicitly discussing whether shutting down fraudulent accounts would hurt revenue too much.
Meta disputes the characterization. Spokesman Andy Stone said the anti fraud team was always meant to be temporary and that Zuckerberg instructed teams to intensify scam enforcement globally. Meta also said it blocked 46 million ads from Chinese partners over the past 18 months and cut ties with bad actors.
Still, the documents suggest the company repeatedly weighed enforcement against the risk of undermining billions of dollars in ad sales.