ChatGPT ban: proxy services blocked on Chinese social media as scrutiny of uncensored AI increases
- Several WeChat accounts offering mainland users access to OpenAI’s ChatGPT have either been suspended or ceased operations
- One operator, whose ChatGPT proxy service remains operational, said they were requested by regulators not to speak to the media
A number of services providing ChatGPT access to mainland Chinese internet users have been blocked, triggering questions over whether the country’s authorities are clamping down on usage of the artificial intelligence (AI) chatbot.
Since San Francisco-based start-up OpenAI launched ChatGPT in November, dozens of public accounts have sprung up on Tencent Holdings’ super app WeChat, embedding ChatGPT via a so-called application programming interface (API) that enables two pieces of software to communicate.
These WeChat accounts had allowed mainland-based users to use ChatGPT, which is officially unavailable in China, without a foreign phone number or a virtual private network. But some of them, including Yibai Technology, ChatGPTRobot, Shenlan BL and AI Duihua, have since been disabled, according to checks by the South China Morning Post on Wednesday.
Yibai Technology has removed a button for activating ChatGPT, while ChatGPTRobot’s mini-program has been suspended for “violating relevant laws and regulations”, according to a notice on the account.
Shenlan BL and AI Duihua told users separately over the weekend that their ChatGPT access services had become “unavailable” without specifying a reason. They asked users to move to new public accounts offering the same services, which were still operational as of Wednesday afternoon.
None of the account operators or Tencent responded to requests for comment.
Some individuals said they have also encountered similar issues.
One user, who requested anonymity due to the sensitivity of the topic, said his personal WeChat account was blocked last week, three days after he embedded a ChatGPT API. He said WeChat had told him he could only regain access to the account after promising not to “arbitrarily” embed APIs in his account again.
Chinese authorities have strengthened investigations into ChatGPT-like services and plan to shut down domestic proxies offering access to ChatGPT, according to a report on Tuesday by local newspaper 21st Century Business Herald that has since been deleted.
Tencent and Ant Group, the financial technology affiliate of Alibaba Group Holding, have been told by Chinese regulators not to offer direct or third-party ChatGPT services on their platforms, according to a report by Nikkei Asia on Wednesday.
DingTalk, the enterprise collaboration platform of Alibaba, has deleted a user guide published earlier this month providing instructions on how to add a ChatGPT bot to group chats.
The Cyberspace Administration of China and the Ministry of Industry and Information Technology, the country’s two major authorities overseeing the internet sector, did not immediately respond to requests for comment by fax on Wednesday.
Two operators whose ChatGPT proxy accounts on WeChat remain operational said they have not received any special notices from the app, although one of the operators said regulators had asked them not to speak to the media.
China has long maintained strict control of the flow of online information from abroad through the Great Firewall, the world’s largest and most sophisticated censorship system that blocks Google, Facebook and other major sites on the mainland. The restrictions have shielded domestic internet giants from competition with global peers, allowing Chinese tech companies to flourish at home.
As ChatGPT gains traction around the globe, some Chinese companies have said they are developing their own rivals to the US chatbot. Baidu, operator of China’s most popular search engine, said it aims to complete internal testing of its Ernie Bot next month before making the chatbot available to the public, while e-commerce giant Alibaba, owner of the Post, also said it is testing a ChatGPT-like service.
