China warns against 'data poisoning' of AI models to distort WWII history

At a regular press briefing on October 9, Chinese Foreign Ministry spokesperson Mao Ning responded to a question about Japanese right-wing forces feeding tampered WWII historical data into international mainstream AI models, causing some models to output distorted historical narratives, such as describing the Marco Polo Bridge Incident as a "military conflict between China and Japan" and downplaying the Nanjing Massacre.

Mao said Japan's militarist aggression is an undeniable fact, and using new technology for data poisoning is a new tactic by Japanese right-wing forces to whitewash and beautify its history of aggression. The aim is to distort global public perception of WWII and remove obstacles for Japan to accelerate its "re-militarization" by breaking free from the post-war system.

Mao stressed that AI should be used for good, not to alter historical truth or absolve aggression. The international community must stay vigilant and jointly resist such acts, which seriously hurt the feelings of people in countries invaded by Japan.