FB Pixel no scriptAlibaba AI Labs to introduce AI cold call terminator | KrASIA

Alibaba AI Labs to introduce AI cold call terminator

Written by Luna Lin Published on   2 mins read

Tired of endless unsolicited sales calls?

Chinese state broadcaster China Central Television (CCTV) revealed how Chinese companies are making billions of unsolicited sales calls by using illegally obtained data and with the help of AI technology. Not long after this, Alibaba AI Labs announced a tool that could help people who feel bothered by these calls. It’s a “cold call terminator” –  an AI chatbot called Erha who will take the call on your behalf.

Erha imitates human responses to incoming sales calls and makes conversation with the salesperson. Alibaba showed an example of how this works in a video that went viral. The product is still in an internal testing stage for now.

“Did you call me before? Your voice sounds familiar,” the chatbot asks a female caller who was promoting bank loans. The responses from the chatbot are so natural that at the end of the 93-second-long video, the still unsuspecting saleswoman asks if she can add the chatbot on WeChat, China’s most popular messaging app.

Junk calls have long been a headache for Chinese mobile users. A company cited in CCTV’s expose admitted that they have made more than 4 billion junk calls in about one year’s time.

13 Chinese government agencies launched a campaign to crack down on junk calls last July, with a focus on unsolicited sales calls.

“People who have watched the video might think this is a very funny product, but in fact, it applies many cutting-edge technologies,” said Nie Zaiqing, the lead developer of the chatbot and principal engineer of Alibaba AI Labs. Erha will automatically responds to calls when incoming caller’s number is already on a blacklist.

In the future, Erha will also be able to make phone calls, take, and leave voice messages, according to Nie. “Think of it as your voice double, and you can use any voice you like to answer phone calls.”


Write to Luna Lin at [email protected]


Auto loading next article...