KATHMANDU: Meta, the parent company of Facebook and Instagram, has said it took action against more than 3.7 million accounts, pages, and other content linked to online scams so far this year, based partly on information received from the Singapore Police Force. The announcement, made during a briefing on platform security, underscores the enormous scale of fraudulent activity on social media and the growing importance of law enforcement cooperation in content moderation. According to Meta, between January and June, it removed more than 113,000 accounts, pages, and other entities related to scams and fraud on Facebook and Instagram. In June alone, the company said it acted against more than 33,600 accounts and pages connected to groups attempting to scam families looking for cheap travel packages during school holidays. Then in July, Meta said it took action against more than 3.6 million accounts and pages linked to scams. Most of these were described as “shell pages,” meaning pages that currently contain no content or are not actively in use. Meta said such pages were removed because they could be purchased or taken over by criminals for future scam activities. The company did not release comparative data from the previous year, so it is unclear whether the numbers reflect a rise in scam activity or an improvement in detection. The disclosure comes as social media companies face increasing pressure from regulators around the world to do more against online fraud, fake accounts, and coordinated bad actors. By highlighting its cooperation with Singapore Police, Meta appears to be demonstrating a commitment to proactive enforcement, but the lack of baseline data leaves many questions unanswered for independent researchers and the public.
According to Meta, cooperation between the Singapore Police Force and Meta on information exchange has been ongoing for the past two years. The company has said that through this cooperation, Meta’s investigators have been able to help identify and dismantle groups involved in scams and the various networks they operate. Daryl Pun, Meta’s Director of Law Enforcement for the Asia-Pacific region, said that cooperation with the Singapore Police helped the company gain information about risks and activities that it could not identify through its own resources. He said the information supplied by Singapore police enabled large-scale action against scam networks, which helped to stop their activities before they reached many people. The partnership is notable because scam operations often span multiple countries and digital platforms, making it difficult for any single entity to see the full extent of the problem. Singapore has been especially active in fighting scams by cooperating with banks, telecom providers, and online platforms. For Meta, the relationship offers actionable intelligence that can guide enforcement priorities, such as which accounts to remove first. For the Singapore Police, the partnership helps disrupt scam infrastructure affecting citizens even when the criminals are based outside the country. Meta has framed this cooperative arrangement as a key part of its broader security system to protect users from fraudulent content before it appears in their feeds. It also reflects a shift in how platforms tackle organized scam networks, moving beyond purely automated content detection and towards shared threat intelligence with government agencies.
The July action accounts for the bulk of the total enforcement figures, with more than 3.6 million accounts and pages removed in a single month. Meta explained that most of these were shell pages, meaning digital properties that appear empty or inactive but could later be acquired or hijacked by criminals. The rationale behind removing such pages is preventive: even if they are harmless now, they can be weaponized in the future to host fraudulent advertisements, phishing links, fake promotions, or impersonation schemes. By removing them before they are activated, Meta hopes to break the supply chain that scam networks rely on to reach potential victims. The company acknowledged that because scam groups operate simultaneously in various countries and on various digital platforms, no single organization has the complete picture of the problem. That statement is significant because it suggests that even millions of removals may represent only a fraction of the global scam economy, which also exists across email, messaging apps, search engines, and other social platforms. The volume also raises questions about how much content is found proactively compared to content reported by users. The June action against travel-related scam pages is a clear example of how scammers adapt to seasonal behavior: as school holidays approached, fraudsters created fake websites and pages offering cheap travel packages to lure families. Meta said it removed more than 33,600 such accounts and pages in June. Still, without previous-year comparisons, it is hard to tell whether these removals are preventing harm or merely keeping pace with new scam content being created every day.
Clara Koh, Meta’s Public Policy Director for Southeast Asia, said scam groups are now operating on an industrial scale. She said they use artificial intelligence to create a large number of fake profiles, as well as fake images and other content, and even produce harmful code. This represents a major escalation from older forms of online fraud, which often relied on manually created accounts or basic phishing emails. With AI, scammers can generate a nearly unlimited supply of realistic-looking identities and material. They can automate the creation of fake profile pictures, fake reviews, fabricated celebrity endorsements, and convincing advertisements. They can also write malicious computer code to build fake websites or applications, further increasing the sophistication of their schemes. AI tools also allow scammers to use automated chat systems to respond to victims in real time, building trust and keeping them engaged over long conversations. This makes it more difficult for ordinary users to distinguish between legitimate and fraudulent pages. For Meta, the widespread adoption of AI by scammers means its own automated systems must constantly evolve to keep pace. The company has invested heavily in machine learning models that detect patterns of abuse, but the same technology is available to bad actors. The result is an arms race between detection and evasion. Koh’s comments highlight the possibility that Meta’s enforcement numbers, while large, are not enough. The industrial-scale production of scam content means that takedown operations can feel like trying to empty the ocean with a bucket. This is why cooperation with law enforcement agencies like the Singapore Police Force is crucial: it enables Meta to target the criminal networks and infrastructure behind the accounts, rather than only reacting to individual posts and profiles.
The announcement comes at a time when online scams have become a top concern for governments in Asia-Pacific and beyond. In many countries, regulators are introducing new rules that require platforms to verify business users, remove fraudulent content promptly, and cooperate with authorities. Meta has said it will continue its efforts to further strengthen security systems on its platforms and to identify and remove scam-related content before users see it. The company’s partnership with the Singapore Police Force is an example of the kind of public-private cooperation that is increasingly seen as essential to combating organized cyber crime. Scam networks often operate across borders, using payment infrastructure, money mules, and digital advertising to reach victims. No single company or country has all the information needed to shut them down completely. By sharing intelligence with Meta, the Singapore Police can trigger removals that protect users beyond Singapore’s borders as well. Meta’s action against shell pages also signals a more proactive approach: instead of waiting for a scam to cause financial damage, the company is removing assets that could be weaponized later. This kind of strategy is common in cyber security, where defenders focus on disrupting infrastructure rather than merely cleaning up incidents. However, the lack of comparative data from previous years remains a weakness. To fully evaluate Meta’s anti-scam efforts, independent researchers and regulators need more detailed transparency reports, including how many reports came from law enforcement, how many accounts were restored after appeals, and how many users may have been harmed before removal. Meta has not yet provided those details. The company has said it is committed to removing scam-related content before users see it, but public confidence will depend on consistent data and measurable outcomes. As scams become more sophisticated, regulators are likely to demand even more collaboration from technology firms.
Looking ahead, Meta has pledged to continue investing in technologies, partnerships, and safety teams to stop scams before they reach users. The removal of more than 3.7 million scam-related accounts, pages, and content items is a considerable enforcement action, but it also highlights how enormous the challenge has become. The use of artificial intelligence by scammers means that platforms, policymakers, and users must all adapt. Meta says it will continue to strengthen security systems on Facebook and Instagram, but the company acknowledges that no single organization has the complete picture of the problem. In the end, the fight against online scams will require coordinated action from technology companies, law enforcement agencies, financial institutions, telecommunications providers, and the public. Meta’s cooperation with the Singapore Police Force has been presented as a model because it has enabled large-scale action against scam networks. Yet the same industrial-scale approach used by scammers demands an equally robust response from defenders. For users, the safest approach is to be cautious about offers that seem too good to be true, such as extremely cheap travel packages during school holidays. Meta said its June takedown specifically targeted groups that exploited this vulnerability. Users should also be wary of pages with no posting history, unsolicited friend requests, and anyone asking for money or personal information. As artificial intelligence makes fake content more convincing, digital literacy becomes a necessary defense. The company has not provided a timeline for new initiatives, but its message is clear: the fight against scams is ongoing. The numbers announced this year could be seen either as progress or as proof of how widespread the problem has become. Without greater transparency and stronger global cooperation, scam networks may continue to find new ways to reach victims. For now, Meta and the Singapore Police are continuing their information exchange, and the platform says it will do more to protect users in the months ahead.
