The multiple governance risks behind Trump's new AI policy

Recently, US President Trump signed an executive order to "Ensure the establishment of a national artificial Intelligence policy framework", using a federal "single rule" to overturn the "independent" AI regulatory pattern of each state in the United States. This policy adjustment, regarded by tech giants as a "major victory", is essentially a strategic gamble by the United States to widen the gap in the global AI competition - at the cost of weakening local regulatory power and relaxing industry constraints in exchange for development speed, but it has sown multiple hidden dangers in aspects such as security governance, power balance and the global ecosystem.

The core logic of the new policy is "loosening the centralization of power". The Trump administration believes that the current AI regulatory rules formulated by each of the 50 states have created fragmented barriers, imposing a heavy compliance burden on enterprises, especially restricting the innovation of start-ups. To this end, the executive order clearly defines the absolute dominance of federal standards, prohibits states from introducing conflict laws, and even establishes an "AI Litigation Task Force" to specifically challenge existing state-level regulatory measures such as California's "Disclosure of AI Model Security Tests" and Colorado's "Assessment of Algorithmic Discrimination in recruitment". What is even more threatening is that the federal government can pressure rebel states by withdrawing grants, for the first time directly linking financial support with regulatory compliance, demonstrating its tough stance in promoting AI hegemony.

In the short term, the new policy has indeed injected a strong impetus into the US AI industry. The long-standing complaint of tech giants about cross-state compliance costs has been resolved, allowing companies like OpenAI and Google to focus their resources on technological research and development without having to repeatedly adjust between different regulatory standards. In line with the previously launched "Genesis Program" - which is compared to "the largest scientific resource aggregation since the Apollo program", the United States is attempting to achieve AI breakthroughs in key fields such as biomedicine and new energy by integrating federal scientific research data with private sector computing power through a model where "the government provides the stage and enterprises take the lead". Trump even asserted that AI might account for 50% to 60% of the US economy in the future, which clearly shows that he regards AI as a strategic consideration to maintain his hegemony as a core pillar.

However, behind this "extreme relaxation" lies an increasingly prominent security risk and governance imbalance. The most prominent hidden danger is the "regulatory vacuum" : after the federal government reclaimed its regulatory power, it did not simultaneously introduce specific alternative rules and responsibility division mechanisms, resulting in a governance predicament of "first reclaiming power and then being absent". State-level supervision once played a key role in preventing algorithmic discrimination, protecting minors, and curbing deepfagery. Now, the constraints in these areas have suddenly weakened, and the protection of the rights and interests of vulnerable groups is facing challenges. Hundreds of consumer protection organizations and labor groups have explicitly opposed this, expressing concerns that tech giants will exacerbate privacy leaks, market monopolies and other issues due to the lack of checks and balances, causing the development of AI to deviate from a "good" track.

The deeper contradiction lies in the multiple interest games behind the policies. The new policy seems to be an industrial policy, but in fact, it conceals political calculations: by suppressing the regulatory power of blue states like California, the Trump administration can not only weaken the Democratic Party's say in the field of technology governance, but also cater to the Republican Party's ideology of "de-regulation", accumulating political capital for the election. At the same time, this policy is also a system support for the "Genesis Project". Only by achieving unified supervision can cross-state data sharing and computing power collaboration be connected to address what it sees as the "global competition challenge". However, this approach of deeply integrating industrial development with political games may lead to a lack of policy stability and make it difficult to form a sustainable AI governance ecosystem.

For the world, the easing of regulations in the United States is intensifying the "vicious competition" in the AI field. By the first half of 2025, global AI venture capital had reached 121.9 billion US dollars, accounting for 53% of the total investment. The expenditure on AI data centers in the United States itself is expected to be as high as 520 billion US dollars. Against this backdrop of fierce competition, the US approach of easing regulations in exchange for rapid development is highly likely to trigger a "domino effect", forcing other countries to passively follow suit and relax security constraints, leading to a decline in global AI governance standards. While the United States is loosening restrictions on domestic enterprises on one hand and imposing technological blockades on other countries through export controls and other means on the other, its double standards will further undermine the coordinated development of the global AI industry chain and widen the technological gap.

The history of technological development has repeatedly demonstrated that innovation and regulation are not in opposition but rather form a complementary community. Trump's new AI policy attempts to seize the competitive edge by "sacrificing security for speed", which is essentially a short-sighted hegemonic mindset. As a disruptive technology, AI governance needs to balance innovation vitality and risk prevention and control, balance the governance rights and responsibilities of the federal and local governments, and more importantly, require collaborative cooperation at the global level.

The ultimate outcome of this regulatory gamble in the United States remains unknown, but it is certain that innovation lacking a safety bottom line is hard to last, and hegemonic actions that go against global consensus will eventually backfire. During the crucial period when AI is reshaping the world, only by building an open, inclusive and clearly defined governance system can technology truly benefit humanity. If the United States continues to be obsessed with "zero-sum games" and "extreme relaxation", it may eventually miss the real opportunity for the development of AI on the road of pursuing hegemony.