'Build, Baby, Build': Inside America’s AI action plan for global dominance

In what it calls a 'new industrial, informational, and intellectual revolution,' the White House on July 23 unveiled America’s AI Action Plan, a 26-page document that lays out the Trump administration’s vision to win the global AI race.

“Today, a new frontier of scientific discovery lies before us, defined by transformative technologies such as artificial intelligence… Breakthroughs in these fields have the potential to reshape the global balance of power, spark entirely new industries, and revolutionize the way we live and work," the document quotes US President Donald Trump as saying.

"As our global competitors race to exploit these technologies, it is a national security imperative for the United States to achieve and maintain unquestioned and unchallenged global technological dominance. To secure our future, we must harness the full power of American innovation,” he added.

The plan is built around three core pillars: accelerating innovation, building AI infrastructure, and ensuring American leadership in international AI diplomacy and security. It likens the current AI race to the space race and promises to achieve global technological dominance.

Accelerate AI innovation

The first pillar focuses on helping the private sector build better AI faster by removing regulatory hurdles and establishing clear priorities.

The plan calls for rolling back what it terms “onerous” regulations enacted under the previous Biden administration and making sure federal money doesn’t go to states that put too many restrictions on AI.

Free speech has been placed at the centre of AI development. The plan says that frontier AI systems must uphold free speech and American values, directing agencies to revise frameworks to remove references to misinformation, DEI, and climate change. Federal procurement guidelines will also be updated to ensure that contracts are given to developers whose systems avoid ideological bias.

A clear preference has been given to open-source and open-weight AI models, which are seen as critical for innovation and academic research. The government will work to democratise access to compute, support spot markets for AI hardware, establish partnerships with tech companies and strengthen the National AI Research Resource (NAIRR).

But it’s not just about building the tech. It’s also about using it. The plan pushes for faster adoption of AI in key industries like healthcare, agriculture, and energy. Defence and intelligence agencies will also be required to assess and adapt their own AI adoption levels regularly.

The plan also focuses on supporting American workers. The government says AI should make jobs better, not take them away. It proposes AI skill integration in education, tax incentives for training, real-time labour market tracking through agencies like the Bureau of Labour Statistics, and retraining for workers whose jobs may change because of automation.

Next-generation manufacturing is another priority. Agencies have been instructed to fund innovations in robotics, drones, and autonomous vehicles using various programmes such as the Small Business Innovation Research programme, the Small Business Technology Transfer programme, research grants, and the Defence Production Act.

AI's role in scientific research is another component, with funding for cloud-enabled labs, incentives for dataset publication, and policies for data sharing across fields like biology and chemistry. A genome-sequencing programme for life on Federal lands is proposed to aid biological model training. The plan also stresses the need for foundational breakthroughs in AI science, including interpretability and robustness.

The government itself is a major target for AI deployment. A Chief AI Officers Council will coordinate efforts across agencies. Procurement systems will be streamlined, government employees will be trained on model use, and departments will be tasked with identifying high-impact AI applications.

The Department of Defence receives special attention, with mandates to establish an AI proving ground, automate priority workflows, and secure emergency compute access during conflicts. Military colleges will also be developed into AI education hubs.

The plan also includes steps to protect AI innovations from cyber and insider threats and prepare the legal system to handle synthetic media.

Build American AI infrastructure

AI needs serious infrastructure, data centres, power, chips, and the second part of the plan is all about building it.

The White House wants to speed up permits for new data centres and factories. It’s also opening up federal land for development and putting strict limits on using foreign or adversary-linked tech in these systems.

America’s power grid, the plan says, isn’t ready for the demands of AI. The plan proposes stabilising existing systems, enhancing transmission efficiency, and prioritising the connection of next-gen power sources like nuclear fusion and geothermal energy.

Another key focus is bringing chip-making back to the US. The plan calls for continued investment in the CHIPS Programme, streamlining policy hurdles, and fast-tracking AI adoption in chip production.

High-security data centres for military and intelligence purposes are on the agenda. New standards will be developed in partnership with national labs and security agencies to protect sensitive workloads from foreign adversaries.

All of this means new jobs and the US government plans to train workers to fill them. The plan suggests launching programmes for electricians, Heating, Ventilation, and Air Conditioning (HVAC) technicians, and others who will build and maintain AI infrastructure.

Cybersecurity is a constant theme in the action plan. The plan suggests the creation of a new AI Information Sharing and Analysis Center to coordinate threat responses across sectors. Agencies will be required to adapt their cyber playbooks to AI-specific threats, and national standards for AI assurance and incident response will be created.

Lead in International AI diplomacy and security

Finally, the third pillar looks beyond US borders.

The plan calls for exporting the full AI tech stack from hardware, models, to standards, to its allies and partner countries. The Department of Commerce will coordinate industry proposals, and agencies like the Export-Import (EXIM) Bank and the State Department will facilitate secure deals.

A key goal is to push back against China’s influence in global tech standards. The plan also suggests that the State and Commerce Departments will actively push US-friendly AI governance frameworks and oppose authoritarian surveillance standards in international diplomatic and standard-setting bodies.

To protect the US' lead in the AI race, the plan proposes stronger enforcement of AI compute export controls, including location verification features on chips, increased global monitoring, and plugging loopholes in the semiconductor supply chain by targeting subsystems that are not yet regulated.

It also calls for closer coordination with allies to prevent other countries from undercutting US controls, and recommends using tools like the Foreign Direct Product Rule and secondary tariffs to ensure broader international compliance.

National security evaluations of frontier AI models will be expanded, particularly around cyber, chemical, biological, and radiological threats. The plan also calls for enhanced scrutiny of foreign AI systems operating in US infrastructure.

Finally, there’s a push to boost biosecurity. Labs that get federal funding will have to use safe tools for DNA synthesis, and agencies will set up systems to detect and stop bad actors from creating harmful substances with AI.