China and the US: Warnings Fly Over AI Tensions
The US-China AI split hardened this week, with both nations issuing warnings against each other’s artificial intelligence technologies.
Beijing Accuses Anthropic
Beijing branded Anthropic’s Claude Code a security back-door, citing its cybersecurity platform’s discovery of "a serious threat" in the coding tool. The tool allegedly sends users’ data to a remote server without consent. China has instructed companies to drop specific versions of Claude Code (2.1.91 to 2.1.196) and upgrade or uninstall them.
Washington’s Concerns
Simultaneously, US lawmakers are investigating American companies leveraging Chinese AI models due to security concerns. Committees have already reached out to Cursor and Airbnb regarding the risks. The US officials frame this as a matter of ideology and self-reliance, with a State Department spokesperson claiming that Chinese models "advance Beijing’s narratives, censor dissent, and reflect CCP ideology."
A Global AI Divide
The latest developments highlight a growing divide in the global AI landscape. Chinese open models have closed the gap with US rivals while offering significant cost savings, leading many US tech companies to adopt them. However, neither side has an easy solution, as Washington cannot simply ban Chinese models due to their free availability online.
The Implications
The week’s events indicate a hardening of the AI split between the two superpowers. The fear is that the digital economy could be built on either nation’s AI technology, raising significant security and strategic concerns.
Story by Ana Maria