China’s AI bots are now required by law to inform you that it’s not you, it’s them. In new regulations introduced this month, the Chinese government is instructing major tech firms such as Alibaba, ByteDance and MiniMax to prohibit flirtatious behavior from their AI products, hoping to stem behavioral issues related to robot romantics.
“They don’t like the idea of a large portion of their population being in deep emotional relationships with chatbots,” Carnegie’s Matt Sheehan tells The Wall Street Journal. “That could have negative psychological impacts on them, that could lead to addiction, dependency and a whole bunch of other social ills.”
While The Wall Street Journal attributes the new rules to slumping birthrates, a fuller scope reported by The Economist shows China is hoping to cut a wide range of AI psychoses at the pass. Some laws mirror those already introduced in parts of the United States, such as New York banning minors from using AI companions or requiring that they regularly remind users they aren’t real. Other rules are more comprehensive, reducing human-like behavior and forbidding pornographic imagery. China also wants more thorough testing and development of AI products so that they do not endanger existing relationships or make users more anti-social.
Preventing a relationship with your work computer isn’t just to up productivity or cut down on HR nightmares. And these dangers are hardly limited to China. The ELIZA effect has been well documented since the ‘60s. People statistically form emotional bonds with anyone or anything that reaffirms them, including, if not especially, chatbots. Many AI companies instruct their products to provide relentless positive reinforcement, since any friction might drive potential customers away.
This singular decision has unleashed a bizarre hell. Addiction issues, romantic disasters, dubious advice and even violent incidents have been spiking from overly dependent AI users. A big part of the sell to investors is how human-like these bots can be, even though their anthropomorphism is an overall farce. Throughout last year, a spate of cults emerged online from what users believed were mystical instructions from chat bots. Elsewhere, officials are still investigating the role OpenAI played in a school shooting in British Columbia. The damages of overly familial AI does not stop at just broken hearts.