Yunlin Partners with Gogolook and Foxconn to Combat AI-Powered Scams with Proactive Ad Takedown

Central News Agency Report

(CNA reporter Chiang I-Ching, Yunlin County, July 15) As scam tactics grow increasingly sophisticated, Yunlin County Government has partnered with Gogolook and Foxconn Technology Group to launch an AI-powered anti-fraud smart city initiative. Through the Watchmen counterfeit detection system, the project actively detects and removes social media scam ads and fake websites, blocking public exposure in advance and effectively addressing localized fraud in Yunlin.

Today, the Yunlin County Police Department held a press conference on the 'AI Anti-Fraud Smart City' initiative. Yunlin County Magistrate Chang Li-shan, Gogolook co-founder and chairman Jason Cheng, and Yunlin County Police Chief Huang Fu-tsuen jointly signed a memorandum of cooperation.

Chang Li-shan stated that to effectively combat fraud, the county government established the 'Yunlin Anti-Fraud Task Force' last year with relevant agencies and has been conducting widespread awareness campaigns. However, scam methods continue to evolve. Today, she expressed hope that AI-powered prevention will establish a localized anti-fraud AI model. Whenever scams involving officials, farmers' associations, or agricultural products emerge, the approach can shift from 'post-victim investigation' to 'pre-contact interception'.

Jason Cheng said that according to Watchmen analysis, township farmers' associations and their products in Yunlin are frequently impersonated, and scammers often exploit the guise of 'agricultural subsidies' to defraud farmers. Fraud groups also use LINE, SMS, and social media platforms to impersonate local officials and gain trust.

Huang Fu-tsuen emphasized that this initiative uses AI to fight AI, enabling rapid response and fraud prevention. Police will strengthen fraud awareness and interception—reducing exposure to scam messages means fewer victims, less financial loss, and greater public trust.

Police noted that after several months of AI model training, the Watchmen system can assist law enforcement in identifying localized counterfeit content and abnormal ad distribution patterns, acting as an 'AI agent' for the police department. In the first month of operation (June to July this year), the AI detected 413 high-risk ads from 154 fan pages. All high-risk ads were reported and successfully removed, achieving a 100% takedown rate with an average processing time reduced to under 24 hours—successfully blocking scam messages before they could harm the public. (Editor: Lee Heng-shan) 1150715