In response to the rapidly evolving tactics used by fraudsters—including the use of AI to mass-produce fraudulent content, the frequent rotation of accounts, and identity impersonation to evade detection—the Administration for Digital Industries (ADI) of the Ministry of Digital Affairs (MODA) is actively working with four major online advertising platforms—Meta, Google, LINE, and TikTok—to enhance their fraud prevention mechanisms. Through real-time intelligence sharing and technology-driven countermeasures, these efforts aim to prevent the dissemination of fraudulent information at its source.
Minister of the Digital Affairs Yi-Jing Lin stated that, in addition to requiring online platforms, in accordance with the law, to implement real-name verification of advertisers and ensure the disclosure of relevant information, the government has established regular consultation and intelligence-sharing mechanisms. Platform operators convert fraud patterns reported by the government into AI model training data and detection rules, thereby transforming known fraud cases into enhanced defensive capabilities.
By leveraging AI to combat AI-powered fraud, these protective mechanisms are shifting from post-incident response toward proactive prevention, enabling fraudulent content to be identified and blocked before it reaches the public.
The ADI explained that, since the promulgation and implementation of the Fraud Crime Hazard Prevention Act, it has continuously guided and required platform operators to enhance their anti-fraud measures in response to evolving fraud patterns. For 2026, the ADI has identified three key areas for improvement: First, platforms are encouraged to leverage AI-based detection technologies to continuously expand their databases of fraud-related keywords, fraudulent promotional content, and associated indicators, enabling the proactive interception of fraudulent advertisements and fake accounts. Second, platforms are to strengthen identity verification and blacklist management for high-risk accounts, while expanding the identification and inclusion of related accounts to prevent their re-entry into the system. Third, platforms are to provide real-time alerts at high-risk points of contact, including private messages, friend requests, and suspicious links, complemented by fraud-awareness campaigns to help the public identify potential risks in a timely manner.
The four major platform operators also outlined their respective anti-fraud measures at the press conference. Meta conducts enhanced due diligence on high-risk advertisers and, in collaboration with MyGoPen and the Taiwan Digital Security Development Association, has launched the “AI Anti-Fraud Real-Time Verification System.” Through regular intelligence sharing with the government, Meta also traces networks associated with the misuse of its services. Google leverages AI to proactively detect and combat fraudulent criminal organizations and provides the public with AI-powered end-user protection features, including “Circle to Search,” Android scam-call detection, and video selfie identity verification for Google Accounts. LINE accelerates the handling of fraudulent activities through industry-wide collaboration and automated takedown mechanisms, while introducing warning features for impersonation of public figures and suspicious links. TikTok combines proactive detection with protective mechanisms, complemented by user reporting, advertising policies, and labels for AI-generated content to reduce users’ exposure to fraudulent content.
The four major platform operators also demonstrated the effectiveness of their proactive anti-fraud measures through concrete results and data. From March 2024 to mid-July 2026, Meta proactively removed 10.9 million scam advertisements in Taiwan and, by mid-July this year, had disabled more than 40,000 accounts used to place fraudulent advertisements. In 2025, Google blocked or removed more than 8.3 billion advertisements that violated its policies worldwide and suspended more than 24.9 million violating accounts. Using AI, Google detected and blocked more than 99% of policy-violating advertisements, helping protect users around the world, including those in Taiwan. From November 2024 to July 2026, LINE took action against 142,006 accounts involved in fraudulent activities in accordance with the Fraud Crime Hazard Prevention Act. Its automated systems processed such cases in less than one hour. In collaboration with the Taiwan Network Information Center (TWNIC) and the three major telecommunications operators, LINE also took down more than 1,500 phishing domains within 24 hours. From the fourth quarter of 2025 through the first quarter of 2026, TikTok reviewed and removed 45,063 scam advertisements based on reports from relevant authorities and took enforcement action against 784 advertisers.
The ADI emphasized that, going forward, under the guidance of the Executive Yuan’s Anti-Fraud Command Center, it will continue to work with platform operators to strengthen and enhance their anti-fraud measures. At the same time, through the regular publication of transparency reports by platform operators, the ADI will ensure that the public has greater visibility into the platforms’ anti-fraud initiatives and their concrete results. The ADI will also urge platforms to ensure compliance with applicable laws and regulations, leverage technology to shorten the duration of fraudulent content online and reduce its reach, and continue safeguarding Taiwan’s digital security environment.