Taiwan's MODA/NICS Real-Time AI Monitoring for Online Fraud and Disinformation

Taiwan's National Center for Cybersecurity Technology (NICS), operating under the Ministry of Digital Affairs (MODA, established August 2022), runs a round-the-clock AI monitoring system that combines natural-language processing, big-data analytics and deepfake-detection modules to scan online advertisements, videos and social posts for fraud and disinformation. As of May 2024, the system was inspecting up to 30,000–40,000 pieces of content per day, flagging an estimated 5,000–10,000 suspected cases per month, with more than 900,000 inspections logged monthly. MODA has said its rapid-response teams aim to issue public counter-narratives or corrections within roughly 60 minutes of a flagged falsehood spreading.
The system's primary documented use case is combating online financial fraud and scams; deepfake and disinformation detection is a secondary but explicitly stated function of the same monitoring pipeline. Independent technology-policy outlets, including GovInsider Asia and Recorded Future News, have corroborated the existence and stated operating scale of the system in interviews with MODA officials, situating it within Taiwan's broader response to a sharp rise in PRC-linked information operations — Taiwan's National Security Bureau separately reported over 2.16 million PRC-linked disinformation items detected in 2024, up 60% from 2023.
No independent technical audit, accuracy rate, or false-positive rate for the AI classifiers has been published, and the government's own figures are self-reported rather than externally verified. Because fraud detection — not disinformation specifically — is the system's most quantified function, its disinformation-countering effectiveness should be read as plausible but only partially evidenced.

Read the full analysis: https://therecord.media/taiwan-using-ai-to-fight-disinformation