Anthropic urges Uncle Sam to kneecap China's AI ambitions before 2028
Jump to main content
Search
REG AD
AI + ML
Anthropic urges Uncle Sam to kneecap China's AI ambitions before 2028
Claude maker warns authoritarian regimes could set the rules unless Washington tightens chip and model controls
Dan Robinson
Dan<br>Robinson
Published<br>fri 15 May 2026 // 14:33 UTC
AI monger Anthropic wants America and its allies to tighten measures<br>aimed at curbing China's AI progress, warning of the consequences if "authoritarian<br>governments" take the lead rather than Uncle Sam.<br>In a lengthy missive<br>posted on its website, the San Francisco-based org says it expects AI to<br>deliver "transformational economic and societal impacts" in the coming years, and<br>whether the transition goes well depends on where the most capable systems are<br>built first.<br>Since the technology is advancing swiftly, democratic<br>countries have only a limited time in which to act, Anthropic believes. The measures<br>it wants to see are nothing new: enforcing tighter export controls on chips<br>used for AI development, such as Nvidia's GPUs, and cutting off access to<br>American AI models.
REG AD
Recent history suggests these controls "have been incredibly<br>successful," it says. But if Chinese researchers are only several months behind the US in AI capabilities, as many experts estimate, how<br>successful can those efforts have been?
REG AD
AI labs in China have only built models that come close to<br>those in America because of their talent and their knack for exploiting<br>loopholes to get around export controls, Anthropic claims, along with distillation<br>attacks that "illicitly extract the innovations of American companies."<br>Many will suspect this is Anthropic's chief motivation in<br>calling for action against China. Back in February, the Claude model maker accused<br>China-based rivals including DeepSeek of using distillation to train their<br>models by siphoning knowledge from Anthropic's own.
MORE CONTEXT
Lawmakers demand great wall to keep advanced chipmaking gear out of China
Europe wants out from under US tech – but first it has to find the exits
China's not thrilled its AI experts want to leave the country
US appears open to reversing some China tech bans
As The Register pointed out at the time, accusing China<br>of copying, while using content created<br>by others to train your own models, shows a staggering lack of self-awareness<br>from the AI industry.<br>Anthropic's sermon also shows blinkered thinking. It implies<br>that China can only advance by riding on America's coattails, and is incapable<br>of innovating. This is despite the shockwaves generated<br>by the release<br>of the DeepSeek R1 model early in 2025, believed to be on a par with the best US models.<br>Numerous reports also indicate that Chinese organizations have made huge<br>strides with domestically<br>developed AI silicon, and Beijing even tried<br>to discourage tech companies in the country from buying and using Nvidia chips.<br>Anthropic sets out two scenarios for what the world could<br>look like in 2028, a date when it expects "transformative AI systems" to have emerged.<br>In the first scenario, America has "successfully defended<br>its compute advantage," and "democracies set the rules and norms around AI." The<br>second has China overtaking the US, leading to AI norms and rules being shaped<br>by authoritarian regimes, with the best models enabling "automated repression<br>at scale."<br>Another problem with Anthropic's plan is that many countries,<br>especially in Europe, view both American and Chinese AI supremacy as a threat to democracy.<br>There is a concerted push in Europe for "digital<br>sovereignty" to minimize reliance on US technology, for example. Others<br>warn it could erode<br>democracy in America itself.
REG AD
Anthropic can draw little comfort from the Trump administration,<br>which has a constantly shifting attitude to China. Export controls were said not to be high on the agenda during the President's trip to Beijing this week, and it was<br>reported that the US<br>has now cleared around 10 Chinese firms to buy Nvidia's second-most<br>powerful AI chip, the H200.
china<br>artificial intelligence<br>anthropic<br>ai + ml<br>ai<br>softwareai_ml<br>nvidia<br>export controls
REG AD
Science
NASA's Psyche mission set for a brief encounter with Mars
There sure are some clever people on Earth
On-Prem
Exploited Exchange Server flaw turns OWA inboxes into script launchpads
Microsoft mitigation may bork inline images, calendar printing while admins wait for a proper patch
ZTE showcases at GSMA M360 LATAM 2026, driving future business model restructuring - AI & network two-way integration
AI-integrated networks can cut costs, boost 5G efficiency, and help regional telcos shift beyond basic connectivity
Patches
Patch time for Cisco SD-WAN admins as vendor drops yet another make-me-admin zero-day
CISA hands feds super-tight deadline for this perfect-10, actively exploited flaw
PaaS + IaaS
AWS to Quick admins: The access control didn't work, but you weren't using it anyway,...