The news: Internal documents reviewed by Reuters show Meta found that close to one-fifth of its 2024 ad revenues from China—or more than $3 billion—came from scams, illegal gambling, explicit content, and other prohibited categories. China-linked advertisers accounted for roughly one-quarter of all abusive ads on Facebook, Instagram, and WhatsApp globally.

Zooming out: Last month, Reuters reported that Meta’s own analysis suggests it earns $16 billion annually from fraudulent ads. The company began an effort to curb fraudulent ads but quickly cancelled those initiatives.

• In early 2024, Meta set up a China-focused fraud unit that, with new detection tools and tougher human review, cut violating China-origin ads from about 19% to 9% of China revenues in six months.
• Later, after an internal “Integrity Strategy pivot,” the team was disbanded, a freeze on new Chinese partners was lifted, and planned anti-scam features were shelved.
• Once enforcement eased, bad ads climbed back to roughly 16% by mid‑2025, and consultants found Meta’s own rules and whitelisting made large-scale fraud easier to run and harder to shut down.

The issue: A number of factors make China a tricky problem for Meta.

• Meta’s China business runs through a multilayer reseller system of major agencies, hundreds of intermediaries, and end advertisers who obscure their identities, making violations hard to trace or enforce.
• Researchers found Chinese authorities rarely intervene when harmful ads target users abroad, reducing deterrence for scammers.
• A 2025 Meta audit of 800 China-based ad accounts uncovered about $28 million in monthly violating spend, three-quarters of which flowed through protected partners. Staff admitted they were unlikely to face penalties due to the revenue at stake.

Why it matters: This pattern has turned Meta’s China fraud problem into a case study in how large platforms can profit from bad actors while publicly emphasizing improvements in ad quality and AI‑driven efficiency.

• Chinese advertisers generated more than $18 billion for Meta in 2024, about 11% of its total revenues per the company’s own data, and major retail clients like Shein and Temu were reportedly spending about $17 million per day combined.
• Internal discussions acknowledged that Meta had effectively chosen to tolerate higher levels of misconduct in China rather than hold ad quality there to the stricter norms applied in other regions.

Key takeaways: Advertisers need to ask harder questions about supply paths, reseller chains, and partner whitelists. Bad actors have the potential to drive advertising costs up for legitimate businesses, and keeping platforms accountable is good business practice.

It’s true that as long as aggregate performance on Meta remains “good enough” compared with alternative channels, the incentives to fully clean up abusive spend are limited, allowing ad quality to erode gradually in the gap between what platforms promise and what they actively enforce.