Meta announced on August 27 the results of its anti-fraud efforts in Taiwan for the first half of 2026. From January to mid-July this year, it proactively removed over 3.1 million scam ads in Taiwan and disabled more than 40,000 ad accounts involved in fraud. Since March 2024, Meta has cumulatively blocked over 10 million scam ads targeting Taiwan. According to the latest statistics from the National Police Agency’s “165 Anti-Fraud Dashboard” under the Ministry of the Interior, telecom and online fraud has shown a continuous downward trend. In July 2026, compared with August 2024, the number of reported cases decreased by 43%, and financial losses decreased by 71%, indicating that the public-private collaborative anti-fraud strategy is taking effect.
Three Core Strategies: From Platform Detection to Cross-Border Law Enforcement
Building on its cooperation with the Ministry of Digital Affairs, the Ministry of the Interior, and the Criminal Investigation Bureau over the past several years, Meta is advancing its anti-fraud efforts across three major fronts.
Strengthen platform fraud prevention measures.Meanwhile, Meta continues to use AI to automatically detect fraudulent content, has introduced celebrity facial recognition technology to prevent impersonation scams, and requires advertisers to complete identity verification and disclose information about ad funders and advertisers. In the first half of 2026, it will go further by implementing advanced verification mechanisms for high-risk advertising accounts, strengthening the review process, and raising the barrier for fraud groups to carry out their operations.[1]
Crack down on scam syndicate networksOn this front, Meta launched a cyber disruption operation in collaboration with law enforcement agencies across multiple countries. One specific case: Meta discovered that a Tainan advertising company was mass-publishing scam ads impersonating investment experts, luring people into communication groups outside Meta’s platform. Meta immediately disabled the related accounts and sent formal notices to preserve evidence, while proactively sharing intelligence with the Taiwanese government. After a coordinated crackdown led by the Fraud Crime Prevention Center, the Tainan District Prosecutors Office, together with the Criminal Investigation Bureau and other units, successfully dismantled this scam-trafficking group operating under the cover of a marketing company—becoming a representative case of public-private cooperation in combating fraud under the “Clean Stream Project.”
In addition, Meta used suspicious information reported by the Taiwanese government as a starting point to proactively expand its investigations on the platform. Beyond blocking fraudulent ads that violated policies, it also removed over 99,000 scam posts from Facebook groups and took down more than 2,300 Threads accounts.
Joining forces with partners to build a defense line.On this front, Meta maintains regular meetings with the Ministry of Digital Affairs and the Criminal Investigation Bureau to exchange information on the latest fraud patterns. Take the recent free-gift and low-price shopping scams in Taiwan as an example: criminals lured people through Threads posts featuring low-priced agricultural products, fan merchandise, and cheap goods. After identifying the latest methods, Meta acted swiftly, and from June 2026 to the present, it has suspended more than 7,900 Threads users.
165 Anti-Fraud Dashboard: Overall fraud declines, but online shopping fraud surges
Although overall fraud data continues to improve,165 Anti-Fraud DashboardThe data also reveals new threats. According to statistics released on August 25, 2026, the number of accepted cases involving online posts and one-page shopping scams increased by 56%, while property losses surged by 733%, making them a new hard-hit area for fraud.
The Executive Yuan’s Anti-Fraud Command Center previously also named Meta-owned social media platforms as a major channel for fraud rings. According to Ministry of the Interior statistics, from early January to the end of April 2026, Meta platforms were involved in 14,073 fraud cases, accounting for 84.31% of all cases, and the financial losses represented 86.2% of the total.
This data presents a contradictory picture: Meta’s anti-fraud efforts are intensifying, but scam operations are simultaneously shifting their tactics. With ad channels largely blocked, one-page shopping sites and social media posts have become the new breakthrough points.
AI fraud detection tools and anti-fraud education
Beyond technical tools, Meta has also invested in anti-fraud education. Last year, Meta donated NT$1 million to support MyGoPen and the Taiwan Digital Security Development Association in creating the “AI Quick Fraud-Detection Assistant.” Users can enter links or messages into Messenger to quickly receive an AI preliminary assessment of fraud risk.
Meta also collaborated with five Taiwanese creators, including Lin Yi-ying, 1+1 Illy, Qiu Yi-shu Joyce, Neinei and Guanguan, and Wen’s Day, weaving anti-fraud concepts into social content in a style close to everyday life. In addition, Meta hosted the “Digital Citizen Summer Tour” in Taipei, Hualien, and Lugang to strengthen the public’s practical anti-fraud capabilities.
Ministry of Digital Affairs Response
Minister of Digital Development Lin Yi-ching stated that combating online fraud requires joint efforts from the government, online platforms, and the public. Meta has cooperated with related anti-fraud efforts, including proactively reporting and removing illegal advertisements, as well as strengthening anti-fraud mechanisms and related features. Lin Yi-ching previously stated at the “National Anti-Fraud Action Forum” that the fight against fraud is “a battle between the good and the bad,” and thanked Meta for joining the “team of good people.”
Rafael Frankel, Meta’s Vice President of Public Policy for Asia Pacific, stated: “Fighting fraud is a battle of speed and adaptability against scam syndicates, and it is also a never-ending task. Meta will continue to invest in cutting-edge technology and deepen collaboration with the Taiwanese government and partners across various sectors to jointly build a safe digital environment for Taiwan’s community.”
Conclusion
Meta’s anti-fraud data reveals the continuously escalating battle between digital platforms and scam operations. Behind the 3.1 million ads taken down and 40,000 accounts disabled lies a combined effort of AI detection, advanced review processes, and cross-border law enforcement collaboration. However, the 165 dashboard simultaneously reveals a 733% surge in one-page shopping scams, demonstrating how rapidly fraud tactics are evolving. As traditional advertising channels get blocked, scammers are shifting to new battlegrounds such as social media posts and low-cost shopping. This fight is far from over. For Taiwanese users, the most practical line of defense remains staying vigilant and leveraging AI-powered fraud detection tools—when encountering suspicious links or low-price product ads, a little more verification means a little less loss.
In addition to cracking down on fraud, Meta should also pay more attention to consumer rights and not use its incompetent AI to lock people’s accounts at the drop of a hat with no avenue for appeal. This situation is even more troublesome for the public than the flood of online scams.
Source: KOCPC Chinese