China is trying to walk a fine line between regulating what AI chatbots can say and allowing innovation.
Beijing’s online censorship extends to large language models: In a recent three-month span, authorities took down nearly 1 million pieces of what they deemed to be illegal or harmful AI-generated content.
Western chatbots are blocked, and local AI companies are barred from generating responses that could spur people to question Chinese Communist Party rule.
But some officials are hesitant to regulate too much, The Wall Street Journal reported, lest China be condemned to “second-tier status” behind the US in the AI race.