Will China’s lead in AI regulation force the US to rethink its approach under Trump?
Trump scrapped an AI model review order over China competition worries, while China already uses a filing system for generative AI services.
Intelligence analysis by GPT-5.4 Mini
The article contrasts the US pulling back on AI safety review plans with China’s more established registry system for AI models and services. It frames that gap as a possible prompt for Washington to rethink how it regulates fast-moving AI.
The story is about two big countries trying to steer the same fast train. One country, the US, is backing away from extra checks on AI tools because it wants to stay ahead of China.
China already has a rulebook. Before many AI products can be released, companies must hand in papers and get them checked, kind of like asking for a stamp before opening a store.
The article says this could make the US think differently. If China already has a system and the US does not, leaders may start asking whether speed is enough or whether some safety rules are needed too.
Analysis
US policy pulls back
The article says President Donald Trump scrapped plans for an executive order that would have directed government agencies to review advanced AI models. His stated reason was to avoid slowing the US lead over China, with the White House treating AI approval requirements as a possible blocker.
China’s system is already in place
By contrast, China has had a registry system since 2023. Developers of generative AI services must report products to the Cyberspace Administration of China, the country’s top internet regulator, before release. The process is layered: documents go first to the provincial CAC and then to the central CAC. The filing package includes a security self-assessment report, a keyword interception list and testing questions. A law firm cited in the article says the procedure usually takes three to six months.
Two tracks for different products
The article also distinguishes between full generative AI services filings and a lighter registration path for some AI applications, such as AI agents that use approved third-party model APIs. That simpler route does not require a security self-assessment report and is reviewed only at the provincial level.
As of April, the CAC had published updates showing 868 filed generative AI services and 530 registered ones. The article’s larger point is that China’s more settled rulebook may pressure the US to reconsider whether speed alone should define its AI strategy.
Key points
- Trump scrapped a planned executive order that would have pushed government review of advanced AI models.
- He said he did not want anything that could weaken the US lead over China.
- China has required generative AI services to go through a filing system since 2023.
- The Chinese process includes security self-assessments, keyword lists and testing questions.
- As of April, China had 868 filed generative AI services and 530 registered ones.
