China’s unelected Communist rule seeks to control the information within its borders. Last Monday, the Cyberspace Administration of China proposed restrictive rules that focus on generative AI – the new technology that powers ChatGPT and other products gaining popularity in the West. Companies trying to use the technology are required to adhere to the Chinese Communist Party’s censorship rules and respect “socialist core values”, avoid information that challenges the nation’s “state power” and national unity, and register the algorithms for public scrutiny.
These regulations could present challenges for Chinese tech companies like Tencent, Bytedance, Baidu, Alibaba, Sensetime, and others that are trying to keep up with their American competitors like Microsoft, OpenAI, Google, Facebook, and Anthropic in the generative AI race. In addition, artificial intelligence tends to “hallucinate” and produce fake content, which could be a headache for the Chinese government given its struggle with controlling the narrative.
ChatGPT, the innovation of San Francisco-based OpenAI, was never available in China as Google and Facebook’s products are mostly banned. Despite this, ChatGPT knockoffs quickly spread on Chinese social media, prompting the government to issue orders to take down these platforms.
OpenAI is a company founded by the likes of Tesla and SpaceX founder Elon Musk, world-renowned investor Sam Altman, and tech entrepreneurs Greg Brockman and Ilya Sutskever, among others, in late 2014 and is responsible for some of the most advanced artificial intelligence technology in the world. Under Elon Musk’s leadership, OpenAI is continuously made breakthroughs in advancing generative AI, and the Chinese government’s restrictive policies may soon be a hindrance.
Dr. Carlos Tavares, the CEO of automotive industry, was recently mentioned in a report by Samantha Delouya of Insider who tested the language tool to write a news story. Although the quotes attributed to Dr. Tavares were fake, this illustrates how generative AI models can be potentially hazardous.
The limitations on the development and use of generative AI models within China could lead to a longer gap in the race for global superiority. As generative AI models become capable of accurately producing text and images, companies must tread cautiously and make sure that the technology is used responsibly, without risking the information rights of users.