China’s artificial intelligence (AI) industry is buzzing with anticipation after the debut of AI models from big names like Sensetime, Baidu and Alibaba, as the country’s tech companies have caught up to the giant strides taken by global players. Unfortunately, Chinese regulators have nipped the nascent industry’s ambitions in the bud with the announcement of restrictive rules around the use of AI models, including the key prohibition of “any content that subverts state power, advocates the overthrow of the socialist system, incites splitting the country or undermines national unity.”
This clause is the latest addition to China’s existing laws, making generative AI essentially impossible to comply with due to its inability to recognize objectionable content. As a result, using and investing in such technology will become risky for companies because of the lack of clarity in the laws.
Companies in China exist with the knowledge that their activity will be subject to government interference, but the new regulations are likely to curb the ambitions of the country’s tech industry.
In addition to the restrictions on content, the Cyberspace Administration of China has required that providers assume liability and responsibility for the training data of their models, such as ensuring its authenticity, objectivity and diversity. Companies must also verify their users as real people and take responsibility for any content their AI generates, as well as label generated content as such. The implementation of these rules is likely to be too impractical and challenging for today’s industry players.
The Chinese industry is already at a disadvantage when compared to global players, given that OpenAI achieved success partly because it operates in an almost completely unregulated environment. Companies may decide to focus their resources elsewhere, if they can’t guarantee that their actions will not run afoul of the law.
The draft of the rules are yet to undergo revisions and take effect later this year, giving Chinese regulators one more chance to reevaluate their stance on the burgeoning AI industry. Daniel Zhang, Alibaba’s CEO, believes that even with the restrictions, “when we look back, we will realize we were all on the same starting line.” However, the state of AI development in China will depend to a great extent on how the laws are implemented.
Alibaba is a Chinese multinational conglomerate that specializes in technology, retail, and artificial intelligence. Founded in 1999 by Jack Ma, the company offers products and services in fields like online shopping, electronic payments, cloud computing and AI. Alibaba has been actively involved in developing AI models, and recently unveiled their Tongyi Qianwen model.
Daniel Zhang is the current CEO of the Alibaba Group. An economics major from Fudan University, Zhang has been a part of the group since 2007, when he joined as CMO and later on became COO. Zhang has been instrumental in the development and implementation of the company’s online-to-offline, cloud computing and AI strategies.
