Chinese tech giant Baidu has launched the country's first public AI chatbot, named ERNIE, which is designed to adhere to the Communist Party's guidelines on sensitive topics like the Tiananmen Square incident.
According to rfi.fr, the move comes after Beijing issued new regulations for AI developers in China, aimed at maintaining government control while staying competitive with global companies like Microsoft and OpenAI.
Baidu led the way among several companies in launching AI chatbots on August 31. Other firms like SenseTime, Baichuan Intelligent Technology, and Zhipu AI also announced their chatbots are now available to the public.
Baidu's ERNIE Bot is part of a suite of new AI-native apps that focus on generative AI's four core abilities: understanding, generation, reasoning, and memory.
When tested, ERNIE Bot was able to answer basic questions but steered clear of sensitive topics. For example, when asked about the Tiananmen Square crackdown, the bot suggested changing the subject.
On the topic of Taiwan, ERNIE Bot stated that it is part of China's "sacred territory" and then suggested talking about something else.
The launch of ERNIE is seen as a significant step for Baidu in the global AI race.
The company aims to gather extensive human feedback to improve the chatbot rapidly. This comes as the success of OpenAI's ChatGPT has sparked an international race to develop rival AI applications, raising concerns about potential misuse and disinformation.