How Meta’s AI tools and partners enabled scam crypto ads on Facebook and Instagram

Ads promising 10% weekly returns ran within 20 minutes and reached over 20,000 users, including in India; Meta stated that it has removed its partner directory and is reviewing the programme

author-image
BestMediaInfo Bureau
New Update
Meta tag
Listen to this article
0.75x1x1.5x
00:00/ 00:00

New Delhi: A Reuters undercover investigation has revealed how Meta’s advertising ecosystem, including its AI-powered tools and agencies listed in its own Partner Directory, can be exploited to place fraudulent cryptocurrency ads on Facebook and Instagram, despite the company’s policies against get-rich-quick schemes.

Reuters reporter Jeff Horwitz said he created ads promoting a bogus investment pitch that promised a 10% weekly return, an annualised rate of more than 14,000%. 

The ads ran for about four days and were shown to more than 20,000 users across the US, Europe, India and Brazil, according to Meta’s reporting dashboard, Reuters said. 

Horwitz said he was able to place the ads with the help of agencies listed by Meta in its Partner Directory as “trusted experts” and “Badged Partners”. He approached Vietnam-based Bluefocus Agency, Hong Kong-based Green Orange and Estonia-based Uproas, and told them upfront he wanted to run banned cryptocurrency ads, Reuters reported. 

The agencies proceeded after charging a higher commission, along with a sign-up fee of $30 or less, payable in cryptocurrency. 

Reuters reported that the ad accounts provided to Horwitz were linked to larger Chinese agency partners visible inside the accounts, including GatherOne and a unit of Beijing-based Cheetah Mobile. Horwitz said he planned to spend about $300 on the test buy. 

The probe also flagged the role of Meta’s automation stack. Reuters reported that Meta’s Advantage+ suggested 10 variants of the ads, including alternate visuals and copy hooks. After publishing, the ads moved from “In Review” to “Active” within 20 minutes, and logged well over 100 clicks, with more than three dozen inquiries from users, Reuters said. Horwitz said he informed respondents that the ads were part of a journalistic test. 

Meta spokesperson Andy Stone told Reuters that company policy prohibits Meta’s Chinese partners, which are meant to serve their home market, from working with foreign resellers. After Reuters shared evidence that some “Badged Partners” explicitly offered to break advertising rules, Meta deleted its partner directory and said it was putting some partners and the programme under review, Stone said.

In a separate Reuters report on Meta’s China-linked ad ecosystem, internal company documents reviewed by the news agency said Meta’s China advertising revenue reached over $18 billion in 2024, and that about 19% of it—more than $3 billion—was tied to ads for scams, illegal gambling, pornography and other banned content. Reuters reported that by mid-2025, banned ads had climbed back to about 16% of Meta’s China revenue. 

Stone said Meta’s automated systems blocked or removed 46 million ads submitted through its Chinese business partners over the past 18 months, usually before users saw them. He said scams were “spiking across the internet”, and Meta was focused on rooting them out through technical measures, disruption of scam networks, and work with partners and law enforcement. 

The Reuters findings come amid wider scrutiny of scam advertising on social platforms in multiple markets. 

AI Instagram Meta Facebook crypto ads
Advertisment