The new American president Donald Trump decree on artificial intelligence (AI) marks a more marked strategic turning point in Washington’s technology policy. This order brings the US national security state closer to companies developing AI models. Its primary goal is to preserve American innovation while ensuring the most powerful AI capabilities support U.S. cyber defense, critical infrastructure, and strategic competition with China. The result could be a more divided AI world order, with Chinese models facing increasing scrutiny over security, data and political alignment.
From risk management to strategic competition
Trump’s order “promoting advanced artificial intelligence innovation and safety,” issued June 2, appears at first glance to be a regulatory document. It establishes a framework for identifying patterns of protected borders, encourages voluntary cooperation between businesses and federal agencies, and strengthens the use of AI in cyber defense.
But this order is more than an attempt to manage the risks associated with new technologies. This is a statement about how Washington now understands AI in the context of Sino-US competition. Advanced AI models are no longer treated simply as commercial products or productivity tools. They are increasingly seen as strategic assets linked to national power.
This marks a clear shift in the direction of US AI governance. Under the Biden administration, AI Policy has placed considerable emphasis on privacy, discrimination, consumer protection, misinformation and model safety. National security concerns were present, but they were accompanied by a broader agenda of social risks.
Trump’s new order prioritizes another concern. It places innovation, cybersecurity and geopolitical competition at the center of AI policy. The logic is close to the approach of Trump’s first term in US AI Initiative 2019. AI leadership depends on research, capital, computing power, talent and standards. From this perspective, excessive regulation risks weakening the very ecosystem that gives the United States its advantage.
Border AI models as national security assets
The order therefore accepts two realities. The first is that border AI models now have national security implications. This is why the National Security Agency and the Cybersecurity and Infrastructure Security Agency play such an important role in this order. The main concern is not the general ethics of AI. The question is whether the most advanced models can improve cyberattack capabilities, expose critical infrastructure to new risks, or be used by hostile actors.
The order directs U.S. agencies to develop a classified evaluation system for advanced AI models with high-level cyber capabilities. It also allows companies to give the government early access to the models before wider release. This is presented as a voluntary process and not a mandatory approval mechanism.
This concept is important. Washington wants an earlier look at the capabilities of frontier models, but it does not want to create a formal licensing regime. The administration appears to believe that formal approval would hurt the speed of AI development. In an industry where progress can happen in a matter of months, delay itself becomes a strategic cost.
This is the second reality underlying this order: an arduous approval system could slow American innovation and strengthen China’s relative position.
This is also where the US approach differs from the European Union’s AI governance model. The EU has placed greater emphasis on risk classification and compliance with legislation. The Trump administration chose a more flexible framework. It seeks to facilitate government access to powerful models while avoiding rules that could slow private sector development.
The result is a special form of cooperation between state and industry. AI companies are not under direct state control, but the government sets up channels to evaluate, influence and use their most advanced systems. Cooperating companies can gain access to federal procurement, defense networks, and trusted partnership agreements with the United States.
Looming competition with China’s AI industry
This could help Washington transform business innovation into national capability and also reshape the AI market. A voluntary rating system may become a practical requirement if investors, corporate clients, and government agencies begin to view it as a mark of trust. Larger companies such as OpenAI, Google and Anthropic are better positioned to manage such relationships. Small businesses and open source developers may find it harder to operate in a market where government validation becomes a business advantage.
The implications for China are significant. The order does not directly name Chinese companies, but its strategic significance is clear. AI plays in the same policy category as semiconductors, telecommunications and critical minerals. This is a technology that Washington considers too important to be left to open market competition alone.
Chinese AI companies have already entered parts of the US digital ecosystem through open source models, development platforms and consumer applications. Some Chinese models are used by developers outside of China. Chinese AI applications in social companionship, image generation and other mainstream services have also reached users in the United States.
These companies are unlikely to benefit from the voluntary testing framework in the same way as American companies. They are not natural candidates for U.S. government contracts. Washington is unlikely to treat them as trusted partners in cyber defense or critical infrastructure either.
But the indirect effect could be substantial. If Chinese companies do not submit models to U.S. safety tests, their refusal could be interpreted by critics in Washington as evidence of risk. They could be accused of hiding backdoors, transferring data to China or aligning with Objectives of the Chinese state. If they participate, they could face tough questions about what information must be shared with U.S. national security agencies and whether such sharing would conflict with Chinese laws on data security, state secrets or technology export controls.
A two-track system in a more divided world?
This suggests a likely two-track system. U.S. and allied models can enter into a trusted ecosystem built around government testing, procurement, and security cooperation. Chinese models could face a separate review, focusing on data, influence, censorship and political alignment.
Control of content can become particularly important. Chinese generative AI services are needed to comply with national rules that include standards of political and ideological content. U.S. officials could use this as evidence that Chinese models are not politically neutral. Testing a model on sensitive political issues can become a way of presenting it as subject to Chinese state influence, even without directly proving data theft or cyber risk.
The longer-term consequence is a more divided AI world order. Sino-American competition will not only be about who makes the best models. It will also be about who sets the standards, who is trusted, who accesses the market and whose models are considered safe for use in critical systems.
Trump’s AI order does not slow AI development. Its objective is to put the development of AI at the service of American strategic power. This is why the order is important beyond US national regulations. This shows that Washington now views AI as a central part of the next phase of technological competition with China.