Anthropic expands AI ban to Chinese-controlled firms

Anthropic, the San Francisco AI company that built the Claude chatbot, blocked Chinese-owned firms and their overseas branches from using its AI services. The company said the step protects U.S. national security and prevents misuse by authoritarian governments.

The new rules were built on earlier bans that had already blocked access to Russia, Iran, and North Korea. Anthropic said Chinese-owned companies, even those running abroad, could still find loopholes to get advanced AI and turn it into tools for military or intelligence use.

Anthropic expands AI ban to Chinese-controlled firms

After blocking access from countries like Russia, Iran, and North Korea, Anthropic now restricts companies or organizations that are more than 50% owned by entities in these regions from using its AI tools. These rules still apply even if those companies are registered and operating outside their home countries.

In the past, firms in authoritarian states would create subsidiaries in other jurisdictions and pretend to be foreign-based businesses while still being controlled by parent companies back home. Anthropic said Chinese companies and other restricted entities would use this loophole to access, analyze, replicate, and adopt sensitive AI models that create direct risks to national security.

In its announcement, the company stressed that a Chinese-owned subsidiary operating in Europe, Southeast Asia, or North America cannot be treated as independent from its parent company. This is because it’s still bound by Chinese law, so the authoritarian government can pressure it to share sensitive information or give access to foreign technology. 

Anthropic sees this as a big risk since these foreign governments could use American technology further to develop projects like advanced surveillance networks and censorship systems. Worse, they could feed the technology into autonomous military drones and AI-guided weapons.

Regulators have raised the alarm about such risks multiple times, and some agencies responded by banning the use of Chinese-developed AI platforms like DeepSeek. This one shocked the global tech sector because the AI platform was well-known for its powerful capabilities. 

For years, Anthropic’s chief executive, Dario Amodei, has urged the U.S. to set tougher restrictions on transferring AI technologies to China. He argues that American companies must limit who can access their products to protect national security instead of waiting for the government to force them to comply. 

Amodei and other policymakers refer to Chinese firms like DeepSeek, Alibaba, Tencent, and ByteDance. They claim these firms have invested heavily in building advanced AI systems and have made quick progress compared to rivals produced in Silicon Valley. They warn that if these foreign companies access Anthropic’s models, they could close the gap and channel that knowledge into military applications that could give their governments a bigger advantage worldwide.

Chinese tech companies face tighter U.S. restrictions

Anthropic’s decision highlights a shift in Silicon Valley’s approach to global security. Traditionally, most tech companies steered clear of foreign policy, but Anthropic has taken proactive measures to support national defense, even at the cost of potential revenue. Rather than waiting for new government regulations, the company is implementing its own standards while urging Washington to strengthen export controls before it’s too late.

Analysts suggest that this strategy will bolster Anthropic’s reputation as a safety-focused AI leader and signal that top-tier AI firms are beginning to see themselves as integral to national defense infrastructure.

While the new policies could cost the company hundreds of millions in revenue, its leadership remains resolute, emphasizing that the risks of misuse outweigh any financial downside. Their approach also supports U.S. policymakers in maintaining America’s technological edge amid rising tensions with China, which impacts critical sectors like semiconductors and quantum computing.

Anthropic appears to have calculated that the long-term advantages of safeguarding its technology and aligning with national interests surpass short-term losses. With a $183 billion valuation, Amazon is a major investor with a client base exceeding 300,000 businesses worldwide. The company’s influence is global. Notably, the number of enterprise accounts generating over $100,000 annually is surging at an impressive pace.

Despite this rapid growth, Anthropic’s leadership insists that safety and responsibility remain at the heart of its mission.

If you're reading this, you’re already ahead. Stay there with our newsletter.