This was announced around September 22–23, 2026, following a media briefing and statements from both Meta and the SPF. It forms part of a roughly two-year information-sharing partnership focused on proactive disruption of scam networks rather than purely reactive removals after user reports.
Breakdown of the enforcement actions
The total includes several waves of activity:
- January–June 2026: SPF referred more than 20,000 accounts, pages, and pieces of content showing scam signals. Meta then investigated interconnected networks and removed or restricted more than 113,000 entities and pages—more than five times the number initially flagged. Most involved investment scams (promises of guaranteed or high returns, often using influencer impersonation).
- June 2026 (pre-school holidays): Enhanced joint enforcement targeted a seasonal spike in activity, acting against more than 33,600 entities linked to deceptive e-commerce, investment scams, and related fraud (e.g., fake low-price offers for concert tickets, food, goods, or travel packages).
- July 2026: The largest portion—over 3.64 million “shell pages.” These appear empty or inactive (no ads or obvious violating content) but function as pre-built infrastructure that scammers can quickly activate, purchase, or take over for campaigns. Meta described this as a newer scam vector identified via SPF signals and removed them before they could be weaponized.
Overall figures are sometimes rounded to “more than 3.7 million” or “nearly 3.8 million.” Meta has also reported removing tens of millions of scam ads in 2026 more broadly (separate from these account/page actions).
Context and broader significance
Scammers increasingly operate at industrial scale, using AI to mass-produce fake profiles, imagery, personalized messages, and malicious code. This allows smaller teams to run large volumes of romance, investment (“pig butchering”-style), e-commerce, and impersonation schemes across platforms and borders. Shell pages represent a shift toward building dormant infrastructure in advance rather than creating assets only when needed.
The Meta–SPF model emphasizes two-way intelligence sharing: police signals help Meta map and dismantle entire ecosystems at platform scale, while Meta data has supported SPF investigations (e.g., arrests related to unlawful remote betting). Officials from both sides highlighted the value of acting on infrastructure before scams reach victims. Daryl Poon (Meta’s Director of Law Enforcement Outreach, APAC) noted that scammers rely on fragmented visibility across organizations, and this partnership counters that. SPF Cyber Command leadership similarly stressed collaboration with private platforms to cut off criminal infrastructure.
This fits into Meta’s wider anti-scam work, which has included large-scale removals of accounts tied to Southeast Asian “scam centers” (often involving forced labor), joint operations with agencies like the FBI, DOJ Scam Center Strike Force, and Thai police, and tools/alerts aimed at users. Singapore has treated scams as a high priority, with ongoing regulatory pressure on platforms (including under laws addressing online harms).
Implications and limitations
Proactive takedowns of dormant assets can disrupt campaigns before activation and raise costs for organized groups. However, scammers adapt quickly, migrate across platforms, and leverage AI, so single-platform or bilateral efforts have limits. Meta has indicated it views the Singapore collaboration as a model to expand regionally. Comparative year-over-year numbers were not provided in the announcements, so it is unclear how much of the volume reflects rising scam infrastructure versus improved detection.
In short, the headline reflects a significant, partnership-driven enforcement effort focused especially on preemptive removal of ready-to-use scam infrastructure on Facebook and Instagram, with the bulk consisting of inactive “shell” pages identified through Singapore police intelligence.
