Chinese internet giant Tencent has taken down a local, Chinese-language chatbot developed with Microsoft’s artificial intelligence by a Beijing-based company after it was less than complimentary about the ruling Chinese Communist Party.
“The group chatbot services are provided by independent third party companies,” Tencent said in a statement. “We are now adjusting the services which will be resumed after improvements.”
Since beginning operation in March, the Baby Q and XiaoBing chatbots had recent given a number of responses unlikely to be welcomed inside the Great Firewall, where government censors are quick to pounce on a long list of banned, politically “sensitive” expressions.
Asked if it would agree with the phrase “Long Live the Communist Party,” the Baby Q bot replied tartly: “Why would I wish long life to such a corrupt regime?”
Later, asked by another user if it loved the ruling party, it returned a curt “No, I don’t.”
Baby Q, which was developed by the Beijing-based company Turing Robot, also appeared to harbor some unauthorized views on patriotism.
“What is a patriot?” asked one user. To which the AI program retorted:
“A patriot is someone who still wants to be Chinese in spite of corrupt officials sending their families and assets overseas, the collusion between government and business, increasing tax revenues and growing oppression of ordinary people.”
Asked if democracy was a good idea, the bot replied: “We must democratize.”
Another user picked up a catchphrase beloved of Chinese President Xi Jinping, asking the bot: “What is your Chinese dream?”
“My Chinese dream,” the bot replied, “is to emigrate to the United States.”
Baby Q was taken offline after these exchanges, with online reports claiming that its programmers had been called in to “drink tea” with the internet police.
A Tencent account-holder who had previously chatted with the bot said he was surprised by the move.
“For a computer to say such things is the natural result of logical processes in its programming, and yet the government regards … it as taboo,” the account-holder said.
Shenzhen-based Internet entrepreneur Zhang Jinjun said the chatbot’s “misbehavior” is likely the result of access to data outside of the government’s tightly controlled online parameters and could have come from people it has chatted with in recent months.
“In the space of communicating with a million people, it is going to start to build its own knowledge base,” Zhang said. “It is quite likely that it will draw conclusions that are similar to universal values [of freedom, rule of law and democracy].”
“It is pretty close to the mainstream of political culture and the direction of history,” he said. “It is highly likely that the bot would form ideas critical of China’s political system when viewed from within its own system of understanding.”
An employee who answered the phone at Tencent’s publicity department declined to comment when contacted by RFA.
Reported by Xin Lin
Translated and edited by Luisetta Mudie.