Anthropic bans Chinese-owned firms from accessing its AI services
3 Min Read
Anthropic’s latest policy targets Chinese-linked firms, raising concerns over the politicization of AI access and technological cooperation.
-
The Anthropic website and mobile phone app are shown in this photo, in New York, July 5, 2024. (AP)
Anthropic, the US artificial intelligence startup behind the Claude chatbot, has announced new restrictions barring Chinese-owned companies and organizations from accessing its technology, tightening its rules on what it calls “authoritarian regions.”
Valued at approximately $183 billion and heavily backed by Amazon, Anthropic has positioned itself as one of the leaders in AI safety and responsible development.
The firm had already restricted access to its commercial services in countries such as China, Russia, the DPRK, and Iran, citing what it perceives as legal and security concerns.
Despite those restrictions, the company said some groups had continued to use its services “through subsidiaries incorporated in other countries.”
Wider context
The updated rules close that gap by prohibiting entities controlled, directly or indirectly, by more than 50 percent by organizations based in restricted jurisdictions.
“Companies subject to control from authoritarian regions like China face legal requirements that can compel them to share data, cooperate with intelligence services, or take other actions that create national security risks,” Anthropic claimed in its statement.
The San Francisco–based company said the change reflects its view that “the safety and security of AI development requires collective commitment to preventing its misuse by authoritarian adversaries.”
Founded in 2021 by former OpenAI executives, Anthropic has grown rapidly. The company said this week it now serves more than 300,000 business customers, with accounts generating over $100,000 annually, nearly seven times higher than a year ago.
China mobilizes state apparatus to accelerate AI push
China is stepping up its technology drive by mobilizing all levels of government to advance artificial intelligence, reflecting President Xi Jinping’s call to leverage the country’s juguo tizhi youshi, a systemic, state-led advantage in concentrating national resources for innovation.
At a recent high-level study session, Xi underscored that Beijing’s centralized model can provide the decisive edge in the global tech race. The directive has already translated into swift action on the ground. When AI glasses-maker Rokid submitted a funding request, a district government wired 3 million yuan ($420,000) into its account within eight minutes.
Local authorities are competing to roll out incentives, from housing subsidies for AI talent to major investments in computing infrastructure. In parallel, the public sector is emerging as a crucial customer base. Rokid’s smart glasses, for instance, are now used in state museums and by utility workers.
In short, China is intensifying its drive to develop a homegrown artificial intelligence ecosystem independent of Western technology, preparing for what it sees as a long-term technological showdown with the United States.
Washington has sought to curb Beijing’s AI ambitions through export controls and other measures designed to restrict access to US funding, expertise, and advanced technologies. While these curbs have had some impact, China is countering with a broad campaign to boost self-reliance in AI, an effort that, if successful, could significantly reduce its exposure to US pressure.
Read next: AI safety: ChatGPT offered bomb recipes and hacking tips
