Chinese organisations are utilising cloud services from Amazon
and its competitors to gain access to advanced US AI chips and capabilities
that they cannot otherwise obtain, according to a Reuters report based on public tender documents.
In
a comprehensive investigation, Reuters revealed how Chinese cloud access to US
AI chips is facilitated through intermediaries. Over 50 tender documents posted
in the past year revealed that at least 11 Chinese entities have sought access
to restricted US technologies or cloud services. Four of these explicitly named
Amazon Web Services (AWS) as a cloud service provider, though accessed through
Chinese intermediaries rather than directly from AWS.
“AWS
complies with all applicable US laws, including trade laws, regarding the
provision of AWS services inside and outside of China,” an AWS spokesperson
told Reuters.
The
report highlights that while the US government has restricted the export of
high-end AI chips to China, providing access to such chips or advanced AI
models through the cloud is not a violation of US regulations. This loophole
has raised concerns among US officials and lawmakers.
One
example cited in the report involves Shenzhen University, which spent 200,000
yuan (£21,925) on an AWS account to access cloud servers powered by Nvidia A100
and H100 chips for an unspecified project. The university obtained this service
via an intermediary, Yunda Technology Ltd Co. Neither Shenzhen University nor
Yunda Technology responded to Reuters’ requests for comment.
The
investigation also revealed that Zhejiang Lab, a research institute developing
its own large language model called GeoGPT, stated in a tender document that it
intended to spend 184,000 yuan to purchase AWS cloud computing services. The
institute claimed that its AI model could not get enough computing power from
homegrown Alibaba cloud services.
Michael
McCaul, chair of the US House of Representatives Foreign Affairs Committee,
told Reuters: “This loophole has
been a concern of mine for years, and we are long overdue to address it.”
In
response to these concerns, the US Commerce Department is tightening rules. A
government spokeswoman told Reuters that they are “seeking additional resources
to strengthen our existing controls that restrict PRC companies from accessing
advanced AI chips through remote access to cloud computing capability.”
The
Commerce Department has also proposed a rule that would require US cloud
computing firms to verify large AI model users and notify authorities when they
use US cloud computing services to train large AI models capable of “malicious
cyber-enabled activity.”
The
study also found that Chinese companies are seeking access to Microsoft’s cloud
services. For example, Sichuan University stated in a tender filing that it was
developing a generative AI platform and would purchase 40 million Microsoft
Azure OpenAI tokens to help with project delivery.
Reuters’
report also indicated that Amazon has provided Chinese businesses with access
to modern AI chips as well as advanced AI models such as Anthropic’s Claude,
which they would not otherwise have had. This was demonstrated by public
postings, tenders, and marketing materials evaluated by the news organisation.
Chu
Ruisong, President of AWS Greater China, stated during a generative AI-themed
conference in Shanghai in May that “Bedrock provides a selection of leading
LLMs, including prominent closed-source models such as Anthropic’s Claude 3.”
The
report overall emphasises the difficulty of regulating access to advanced
computing resources in an increasingly interconnected global technological
ecosystem. It focuses on the intricate relationship between US export laws,
cloud service providers, and Chinese enterprises looking to improve their AI
capabilities.
As
the US government works to close this gap, the scenario raises concerns about
the efficacy of present export controls and the potential need for more comprehensive
laws that cover cloud-based access to banned technologies.
The
findings of this paper are likely to feed ongoing discussions about technology
transfer, national security, and the global AI race. As politicians and
industry leaders analyse these findings, they may spark fresh discussions about
how to balance technological cooperation with national security concerns in an
era of rapid AI growth.
