Microsoft’s Chinese-language AI chat bot filters certain topics, the company confirmed Monday, although it did not clarify whether that included interactions deemed politically sensitive.
Last week, CNNMoney and China Digital Times reported that Xiaoice would not directly respond to questions surrounding topics deemed sensitive by the Chinese state. References to the Tiananmen Square massacre of 1989 or “Steamed Bun Xi,” a nickname of Chinese President Xi Jinping, would draw evasive answers or non sequiturs from the chatbot, according to the report.
“Am I stupid? Once I answer you’d take a screengrab,” read one answer to a question that contained the words “topple the Communist Party.”
Even the mention of Donald Trump, the American president-elect, drew an evasive response from the chatbot, according to reports. “I don’t want to talk about it,” Xiaoice said, reports CNN Money.
June 5th: The AI Audit in NYC
Join us next week in NYC to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.
In response to inquiries from Fortune, Microsoft confirmed that there was some filtering around Xiaoice’s interaction.
“We are committed to creating the best experience for everyone chatting with Xiaoice,” a Microsoft spokesperson tells Fortune. “With this in mind, we have implemented filtering on a range of topics.” The tech giant did not further elaborate to which specific topics the filtering applied.
China scored dead last in a Freedom House survey on Internet freedoms last year. Content deemed politically sensitive by the state would be censored, and websites like Google and Wikipedia were routinely blocked.
Microsoft says that Xiaoice engages in conversations with over 40 million Chinese users on social media platform like Weibo and WeChat, reports CNNMoney — a runaway success compared with Tay, Xiaoice’s ill-fated counterpart, which was essentially pranked into sending hateful messages while it was briefly active on Twitter. Fortune was unable to put Xiaoice to the test.
Last week, reports emerged that Facebook might be developing censorship tools in a bid to return to China, which has blocked the service alongside other platforms like Twitter since 2009.
Facebook told Fortune in a statement that it had “not made any decision on our approach to China,” stressing that its “focus right now is on helping Chinese businesses and developers expand to new markets outside China by using our ad platform.”
This story originally appeared on Fortune.com. Copyright 2016
