The US wants cloud service providers like Amazon and Microsoft to disclose foreign clients developing AI applications in China.
The Biden administration’s new proposal requires firms to disclose foreign customers’ names and IP addresses and proactively investigate and report suspicious activity. This initiative is part of a broader tech dispute between the US and China, as both countries compete in the ‘AI race.’ The White House is concerned about Chinese companies access to computing power through cloud providers, notably to train powerful AI models.
Why does it matter?
The proposed draft rule requiring US cloud providers who supply computer capacity for foreign AI training to declare that they do so was unveiled on 29 January as part of “key actions following President Biden’s landmark executive order.” If the Department of Commerce’s plan is finalized and implemented, US-based cloud providers must notify the government when foreign clients train the most powerful models, which might be exploited for adversarial intentions.
The rule aims to thwart “malicious cyber-enabled activities” and limit Chinese firms’ access to crucial data centers and servers for AI applications. The US also investigates ways to force cloud companies to reveal every time a non-US entity taps cloud services for AI development. This move reflects the intensifying geopolitical rivalry between the US and China in the AI and emerging tech domains, especially for dual-use technologies.
On their side, the Chinese government has its regulations related to cloud services, such as the “Multi-Level Protection Scheme 2.0” and the “China Cybersecurity Law,” which impose requirements on using cloud services and data protection. Additionally, Beijing has wide discretion to determine, on a case-by-case basis, whether a network security review is required for cloud services, depending on the impact of such services on national security. The US-China AI competition is a central focus, with both countries vying for a strategic edge. China aims to develop state-of-the-art AI capabilities, including thwarting US export restrictions on the much-needed advanced AI chips. The US is taking various actions to mitigate strategic risks and restrict China’s military AI capabilities while advancing its own. This move shows that, as part of the broader tech and geopolitical rivalry between the US and China, the Biden-Harris administration is pushing for increased transparency from US-based cloud service providers to identify and report foreign clients developing AI applications in China. This is a testament to the growing importance of AI in the strategic competition between the two superpowers.