News - China Tightens Grip on AI by Restricting Access to Foreign Models

The Chinese government has initiated a crackdown on domestic technology providers offering access to foreign artificial intelligence models including ChatGPT and Claude. This move signifies an intensifying effort by Beijing to control AI services within the country. Recently, several AI relay platforms in China that facilitated access to foreign AI models have been forced to suspend or scale back operations. This action follows reports of law enforcement questioning platform operators. Informal communications indicate that authorities in Hangzhou ordered the shutdown of services functioning without the necessary permits. A significant AI relay platform communicated through a group message that it was pressured by Hangzhou police to cease operations and refund subscribers as its services lacked official certification. The platform soon ceased operations following this directive. Because global AI companies like OpenAI and Anthropic do not operate within China, domestic businesses have created relay platforms allowing users to engage with foreign AI tools without creating overseas accounts. Recent developments reflect concerns from China's Ministry of State Security regarding these platforms, citing potential national security issues due to the risk of uncontrolled data transfers. Despite high demand for foreign AI model access, Chinese regulations have made official registration mandatory for generative AI products. Moreover, all data processing must comply with oversight requirements. By April's end, a significant number of generative AI providers were registered, with numerous applications and services using foreign models completing additional filings. Various experts, under anonymity for safety, criticized the Chinese Communist Party's approach, suggesting it demonstrates unease with foreign AI models bypassing Chinese scrutiny and data protocols. Comprehensive restrictions are perceived to disconnect users from international information landscapes, limiting diverse perspectives. The intensified policing has broader implications, particularly among software developers and entrepreneurs who depend on AI relay platforms for back-end processing and product development. Disruptions arising from platform closures complicate these operations and drive businesses toward riskier, less visible alternatives. With Beijing's stated ambition to promote AI innovation, experts argue that stringent restrictions on access to foreign AI might inadvertently hinder progress. They predict that users will transition to smaller, less regulated platforms, increasing operational risk and complicating technology access. Ultimately, the tightened control over AI access looks set to restrain both technological advancement and information plurality within China.