Many global tech companies such as Microsoft, Google, and Baidu have raced to develop their chatbot, a program that can mimic human dialogue using AI.

That of Alibaba is currently being tested by employees, a spokeswoman for the group told AFP, without specifying the launch date or the platform on which it will be available.

This announcement comes a few days after Baidu, Chinese internet giant, said that the testing phase of its chatbot, also working with artificial intelligence and called "Ernie", will be completed in March.

For its part, Microsoft has signed a multi-billion dollar agreement with the ChatGPT publisher start-up, OpenAI, in order to integrate new conversational AI technologies into its Bing search engine.

Google also announced a battery of features based on artificial intelligence on Wednesday.

But these synthetic text, image and sound generation capabilities, if left unchecked, can open the door to identity theft, financial fraud or reputational attacks.

This technology worries worldwide.

The phenomenon of "deepfake", for example, which also uses artificial intelligence to generate hyperrealistic digital tricks, sometimes with the image of celebrities, presents "a danger to national security and social stability", according to Beijing.

© 2023 AFP