US to Limit American AI Models in China
In a bid to curb the dominance of foreign-made artificial intelligence (AI) models, particularly from countries like China and Russia, the Biden administration is reportedly planning to impose limitations on American-made AI technologies.
China's Growing AI Industry
China has made significant strides in developing its own generative AI industry over the past year, urging its companies to reduce reliance on foreign technology. However, questions arise regarding China's current dependence on U.S. AI models and the potential impact of new American policies on Chinese AI development.
OpenAI's Presence in China
OpenAI's renowned AI services like ChatGPT and DALL-E have not been officially accessible in mainland China due to unspecified "local conditions," as stated by an OpenAI spokesperson last year. Despite this, many Chinese entities have managed to utilize OpenAI's services using tools like virtual private networks (VPNs).
Tensions with OpenAI
OpenAI has restricted access to its services for Chinese companies, as exemplified by its suspension of ByteDance's account, the parent company of TikTok, in December. The move came after allegations that ByteDance utilized OpenAI's technology for its AI development.
Open Source vs. Closed Source AI
The U.S. government's proposed restrictions primarily target the export of closed-source AI models. China, however, has heavily relied on open source models developed in Western countries, such as Meta Platforms' "Llama" series, utilized by the Beijing Academy of Artificial Intelligence and other Chinese tech giants.
Quest for Self-Sufficiency
Chinese authorities emphasize the importance of developing "controllable" AI technology, fueling discussions on achieving self-sufficiency in AI development. China has also implemented regulations requiring government approval for the release of AI services to the public, with a focus on domestic models over foreign ones.
