China warns of US AI: Threatens regime security

Recently, warnings from senior technology executives in the US about the rapid development of AI have been pouring in. A renegade scientist at the American company Anthropic went so far as to talk about the risks of the end of humanity unless someone pulls the handbrake.

At the same time, the competition to lead the development of the new technology is razor sharp. Whoever wins the AI ​​race, wins, Donald Trump has said, describing the technology as the oil of the future.

So far, the United States is first, according to most evaluators. But China is not far behind. Big money has been invested in making China the world leader in AI, which is the new battleground between the superpowers.

But also in China, a concern about what the technology can achieve is beginning to be felt. AI poses a threat to the security of the Communist Party, writes Chen Yixin, head of China's Ministry of Security, in a state-run newspaper on digital technology. Among other things, he warns that AI technology can be used by "hostile forces" to "fabricate political rumors and spread harmful information". It threatens China's "ideological security", according to Chen. That is, the one-party rule of the Communist Party.

Chen also warns that US AI models - he singles out Claude Mythos and GPT-5.5 Cyber ​​in particular - could be used for "large-scale espionage" and to attack China's infrastructure.

Another risk for the Communist Party is that AI could generate fake videos, voice recordings and articles that degenerate into a propaganda war against China. Foreign AI models can hack into Chinese social media and take control of millions of users' accounts. Fake videos of China's leader Xi Jinping could spread messages that contradict what he has actually said and thus threaten the Communist Party.

Chen also sees a risk of sensitive information leaking abroad when Chinese use foreign AI models. He notes that Chinese startup Moonshoot AI forwards questions from users to Claude, which is run by American Anthropic. Sometimes the questions have contained sensitive information, including videos from surveillance cameras linked to the Chinese military.

To minimize the risks, Chen emphasizes the importance of China being self-sufficient in the technology required for AI. He also calls for international cooperation on laws and ethical guidelines that reduce security risks with AI.

At the same time, China's rule is critical of the increasingly loud voices of tech giants in the US calling for a global slowdown in AI development, but in a way that prevents China from overtaking the US. Democracies must lead AI's progress, writes Dario Amodei, CEO of the AI ​​company Anthropic, in a well-publicized open letter on his website. If China takes the lead, it opens the door for the Chinese Communist Party to militarily dominate democracies, Amodei warns.

That kind of reasoning is dismissed by Guo Jiakun, a spokesman for China's foreign ministry, as scaremongering that undermines efforts to bring about global AI rules.

- All parties should work together to develop AI in an open, inclusive and mutually beneficial way, he said at a press conference.

The question of how AI should be regulated will be a topic of discussion when China's leader Xi Jinping visits US President Donald Trump in Washington next week. This is where the fight between the countries stands now.