Anti-Fraud Center

According to news reports on 19 May, the National Anti-Fraud Center App has recently officially launched the "AI Content Identification" function, which supports the use of AI-generated traces of images, videos, audio and text, mainly to identify fraud and forgery in the electronics business scene. There is a growing tendency on the electrician platform to use AI to tamper with pictures to obtain "refunds only". Following the random selection by journalists of 10 suspected AI-generated post-sales maps, which were uploaded to the National Anti-Fraud Center App and several AI-based forensic tools, journalists were convicted of "images with AI-generated traces". I don't know

On March 13th, the National Anti-Fraud Center App was upgraded and the AI content recognition function was added. 1AI has measured that the AI content validation function of the National Anti-Fraud Centre supports the AI content testing of images, videos, text and audio. When used, specific content is uploaded according to the corresponding module. Of these, pictures limit the size of 30KB~5MB, video limit 100KB~100MB, text limit 10 words ~5000 words, and audio limit the human voice of less than 10 minutes. In addition, the function currently limits the number of daily tests to 1..