As Meta’s war on scam ads intensifies, the social media giant said that in the past year it has seen a substantial decline in the volume of scam ads that are being reported by users, but the battle is far from won.
At a scam ads briefing to journalists in the Asia-Pacific region, Meta laid out the progress it is making to identify and remove scam ads on its platforms.
“We have the criminal scam syndicates operating out of a number of places in Southeast Asia, where they are driving some of the most adversarial, malicious and criminal scam behaviour we see in the world,” Meta director and head of security policy Nathaniel Gleicher said.
“It’s not just that these scam centres have annual illicit revenues of $40 billion, it’s not just that they’re driven largely by organised crime and criminal gangs, we are seeing them leverage forced labour. The only thing we can call it is modern slavery.
“More than 300,000 people are being forced to run these scams, meaning that every time there’s a scam from one of these scam centres, there are two victims.”
Meta has set up focused teams to target specific scam ad operations across South East Asia, and this year has removed about 12 million accounts, links to those scam centres across Facebook, across Instagram and WhatsApp.
It has also cracked down on user accounts that have been compromised by scammers; in the past quarter there has been a 50 per cent reduction in detected accounts compromised on Facebook and more than a 40 per cent reduction in detected compromise on Instagram.
Meta is working with other Big Tech companies, governments and law enforcement to tackle the issue.
Scam Ads Down
Meta uses a metric called Scam Reports per Million Views (SRMV) to measure the prevalence of scam ads on the platform.
Between July 2024 and June 2025, Meta reported a 59 per cent reduction in the volume of scam ads reported globally, down from 4.5 per million to 1.8 per million, with a similar reduction in the Asia-Pacific region.
In 2024, 157 million ads were removed on Facebook and Instagram for violating Meta’s Frauds & Scams policies with 90 per cent detected before being reported by users.
Reports of scam ads that use a celebrities’ image, known as celeb-bait ads, have seen a lower reduction, down 21 per cent in the APAC region. These account for a much smaller proportion of views with a scam report in just under one in 5 million views.
Scam ads have previously targeted high profile Australians on Meta platforms, including mining moguls Andrew Forrest and Gina Rinehart, footballer Sam Kerr and TV personalities Eddie McGuire, Allison Langdon, Karl Stefanovic and David Koch.
In the past year Meta has been trialling face recognition technology to detect and block scam ads that leverage the faces of celebrities and public figures. It now has 500,000 public figures that have been detected using facial recognition, and expects the detection of celeb-bait ads to markedly improve.
“In June, July of this year, we ran another test, and that showed that we automatically blocked 100 per cent more celebrity scam ads with the facial recognition technology, that’s an 80 per cent improvement (on the past year’s SRMV),” Meta director of product Maxime Prades said.
“Today, we’re experimenting to counter other impersonation threats on Facebook and Instagram with the same technology and continuously iterating and trying to improve on it.”
Aussie Crackdown
In 2024, the Australian government released the Scams Prevention Framework, which established principles that require organisations including social media companies, banks and telcos to prevent, detect, report, disrupt and respond to scam ads.
Organisations that ignore the new laws face fines of up to $50 million.
When asked about whether Meta would comply and how it is preparing for potential compensation claims from scam victims, Gleicher said that Meta hasn’t seen the details of the legislation.
“I want to see those to think through exactly how it’s going to work,” he added. “I will say that, in my experience, cross industry collaboration is the most effective way to counter these scammers and to tackle scams. So I look forward to working closely with our partners in industry and in government and in law enforcement to counter these threat actors.
“We have no interest in having scams on our platform, if we want to build a community where people are going to come and spend time, build with their friends, connect with their community, they need to know who they’re talking to and they need to trust the community they’re talking to.”
The tougher advertiser verification tools being used to weed out scam ads are slowly being expanded to other advertisers, but Gleicher said rolling it out too quickly or widely could prove counterproductive.
“We have to be very careful, because verification is a useful tool but it is not a silver bullet,” he said. “Every time we expand verification, we see scammers immediately try to defeat that verification, which means that by expanding steadily, we can improve the verification and counter scammers that are trying to abuse it.”
