China, the first among major economies to put AI guidelines into law, approved Baidu's Ernie Bot to be open to the public, as the country is walking a tightrope between regulation and innovation for generative AI that may disrupt human life for years to come.
According to Chinese media Sohu and IT Home, China granted Ernie Bot to open for public use, and users can download it from Ernie Bot's website, Apple's App Store, and other app stores on Android in China.
Ernie Bot was launched on March 16 on invitation-based testing, and Baidu has upgraded the AI model to version 3.5. Robin Li, founder of Baidu, said that with the opening of Ernie Bot to hundreds of millions of users, it will gain access to a substantial amount of real-world human feedback, further enhancing the model for future iterations and improvement.
Ernie Bot is the first AI model to be approved by China after seven Chinese authorities, including the Cyberspace Administration of China, unveiled the first AI guidelines in July, which took effect on August 15. It remains to be seen if China will approve other Chinese-developed AI models. In addition to Baidu, Alibaba, Huawei, and SenseTime have showcased their AI models recently.
Although Chinese companies are lagging behind global frontrunners like OpenAI, Microsoft, and Google in terms of generative AI, the Chinese government moves quicker than its peers regarding AI regulations. The US is still holding hearings to collect more comments from AI companies, and the European Parliament passed a draft of the AI Act, which would impose new guardrails and transparency requirements for AI in June.
Bloomberg quoted Matt Sheehan, a fellow at the Carnegie Endowment for International Peace, saying that China started building the regulatory tools and muscles so that they are going to be more ready to regulate more complex applications of AI.
