China moves to develop its own version of ChatGPT amid censorship concerns

Chinese search engine giant Baidu is moving to set up its own artificial intelligence chatbot amid reports that regulators have warned major tech companies not to offer the Microsoft-backed artificial intelligence bot ChatGPT to the public.

Baidu announced on Feb. 22 it will begin incorporating its artificial intelligence chatbot Ernie into its search services from March, as part of an apparent bid to put a stop to the use of ChatGPT-backed services in China.

Baidu CEO Robin Li said in an internal memo that Ernie Bot will be integrated across all of Baidu’s operations, including its search and cloud services, the Associated Press reported, adding that Baidu also plans to integrate Ernie into its smart car operating system and smart speaker.

“AI technology has reached a tipping point and all industries will inevitably go through transformation,” Li"s memo said.

“Baidu stands as the best example of the long-term growth of China’s AI market and is advancing at the forefront of this new wave,” he said.

Fellow tech giant Alibaba confirmed on Feb. 8 that it is also developing a ChatGPT style AI tool, which is currently under internal testing, the state-backed English-language China Daily reported on Feb. 17, without giving a timeframe for its launch.
Lack of control
There have been signs that the ruling Chinese Communist Party is anxious to stop the widespread adoption of an AI bot – which generates text and computer code based on huge "corpuses" of existing text – that operates outside of the government’s strict censorship and "public opinion management" regime.

The government apparently fears that text generated from corpuses it doesn’t control could start appearing on Chinese social media.

The China Daily warned on its Weibo account on Monday that ChatGPT "could provide a helping hand to the U.S. government in its spread of disinformation and its manipulation of global narratives for its own geopolitical interests."
A response in Chinese by ChatGPT, an AI chatbot developed by OpenAI, is seen on its website, Feb. 9, 2023. Credit: Reuters illustration

Last week, Japan"s Nikkei news service reported that tech giants Tencent and Ant Group have been told not to use ChatGPT services on their platforms, either directly or indirectly.

Tech companies will also need to report to regulators before they launch their own ChatGPT-like services, the agency quoted people familiar with the matter as saying.

ChatGPT isn"t officially available in China, although some users were able to access it using tools to circumvent the Great Firewall of internet censorship, and some companies have claimed to offer it as a third-party application on Tencent"s WeChat platform.

Those services have now been suspended, both the third-party ChatGPT services and the Chinese-made copycats, Nikkei reported.

‘Deeply deformed products’

Beijing-based current affairs commentator Si Ling said China"s artificial intelligence developers don"t currently work in a free environment, so the products they create will inevitably be subject to government censorship.

"China wants to compete with the United States in the field of AI, but its lack of talent, mechanisms to promote innovation and freedom of expression will ensure that it can only develop deformed products bearing the hallmarks of the party-state," Si said. 

"[Either that] or steal advanced technologies from abroad."

According to Si, China wants to develop its own AI to step up monitoring of its citizens" daily activities, both on- and offline.
China wants to develop its own AI to step up monitoring of its citizens" daily activities, both on- and offline, says Beijing-based current affairs commentator Si Ling. Here, police officers display their AI-powered smart glasses in Luoyang, Henan province, China, in 2018. Credit: Reuters

The State of AI Report 2022 agreed, saying that Chinese AI-related research papers "focus more on surveillance-related tasks."

"These include autonomy, object detection, tracking, scene understanding, action and speaker recognition," the report found.

China has been putting out AI-related research papers at a furious pace in recent years, it said, with Chinese institutions authoring 4.5 times the number of papers that U.S. institutions did between 2010 and 2022.

AI capital

Beijing"s municipal government announced on Feb. 13 that it will support leading AI companies to build their own versions of ChatGPT, as well as boosting the supply of AI data.

The government said in a white paper on the industry that the Chinese capital is currently home to more than 1,000 top AI companies, accounting for nearly 30% of the nationwide total.

"One of the driving forces behind Beijing"s continued AI lead is the steady influx of technical talents, with over 40,000 individuals accounting for 60% of the country"s total," the English-language Global Times reported.

"Beijing plans to further increase the introduction and training of technical talents, encourage them to start their own businesses in the city, and support existing leading enterprises to further develop innovative applications in autonomous driving, intelligent manufacturing and smart cities," the paper said.

Translated by Luisetta Mudie. Edited by Malcolm Foster.


[圖擷取自網路，如有疑問請私訊]