America First AI Agenda

A vision for American Innovation

The America First Policy Institute (AFPI) introduces our emerging technology portfolio with the America First Artificial Intelligence (AI) Agenda. The America First position on AI is simple: ensure that America leads the world in developing and adopting AI for human flourishing. This document, written by a team of experts with extensive AI policy and industry experience, describes that position.
The America First approach to AI is holistic: balancing the demand for innovation with the duty to uphold American values and human dignity. Striking that balance is critical for unleashing a Golden Age of economic prosperity that puts American values and the American people first.

“I am not myself apt to be alarmed at innovations recommended by reason. That dread belongs to those whose interests or prejudices shrink from the advance of truth and science.”

– Thomas Jefferson

AI is a fast-changing and far-reaching technology. Many AI capabilities have recently progressed more quickly than experts predicted prior to the release of ChatGPT (Kučinskas et al., 2025). In 2025, for example, AI systems achieved gold-medal performance at the International Math Olympiad (Luong & Lockhart, 2025) and generated experimentally validated research on leukemia (Gottweis et al., 2025). These breakthroughs surprised even the most optimistic experts. The AI boom is also translating into real-world economic performance: AI capital expenditures added over a full percentage point to U.S. GDP growth in the first half of 2025 (Aliaga, 2025). If these trends continue, AI may soon become as critical as electricity for a nation’s future success.

Critically, AI has the potential to radically transform life for Americans. In healthcare, for example, AI is accelerating drug discovery and holds the promise to help cure diseases that cut American lives short (Serrano, 2024). In manufacturing, AI is bringing industry back within our shores, creating good-paying American jobs (Thomas, 2025). In the military, AI improves targeting and makes American warfighters more lethal, keeping the American people safer from foreign adversaries (Department of the Air Force, 2025). AI helps farmers grow more food by predicting crop yields and optimizing irrigation (Jabed & Azmi Murad, 2024), makes our power grids more reliable by forecasting outages (Yao et al., 2023), and personalizes education for our children by adapting to students’ individual pace of learning (Létourneau et al., 2025). This technology will determine which nations prosper and which fall behind.

Given AI’s transformative potential, we must ensure that America, not China, leads in AI. We cannot take AI progress for granted. Maintaining our current pace of AI development requires innovation-friendly regulatory environments and resource abundance. For example, AI model development and deployment require exponentially larger amounts of computational power (“compute”), energy, and data (Sevilla et al., 2024). Unfortunately, a dense thicket of regulations and resource constraints holds back American AI capabilities, while China forges ahead. Unless we chart a new course, China could lead in AI and control the 21st century. In that world, America could become completely dependent on the Chinese Communist Party (CCP) for AI systems that are crucial for economic innovation and national security.

However, in the pursuit of AI leadership, America must not ignore its values. In the process of beating China, we cannot become China by allowing AI to enable the censorship, mass surveillance, and authoritarian power consolidation that characterize authoritarian regimes. Nor can we resign our values to those of Silicon Valley, whose systems propagate anti-conservative bias (Mazeika et al., 2025; Rozado, 2023) and willfully output disgusting, pornographic content to children (Knutsson, 2025; Fox News Channel, 2025). When the interests of Big Tech and the American people are not aligned, we must put the American people first.

AN AMERICA FIRST AI AGENDA IS ESSENTIAL

While AI as a research endeavor has existed since the dawn of the computing age, AI as a general-purpose technology is new, and America has not yet developed a comprehensive framework for governing it. In this vacuum, the policy debate has fractured into simplistic approaches that, in advocating for legitimate concerns, jeopardize the well-being of the American people.

The first failing approach is pushing for onerous regulations that would strangle the American AI industry. History shows us why this fails. We need to look no further than the environmentalist movement, which devastated American nuclear energy dominance in the 1970s, slowed economic growth, and convinced a generation of young people that climate change would end the world. We cannot repeat these mistakes by applying the same reflexive, anti-growth instincts to a promising area like AI. If heavy-handed regulation compromises American AI leadership, the 21st century could instead be dominated by the CCP.

The other failing approach is blind optimism: pushing for AI acceleration without any guardrails. Widespread AI diffusion and adoption is indeed critical to American leadership, but it cannot occur until AI is worthy of the public’s trust. The AI revolution could, if unchecked, cause great harm: mass dissemination of pornographic “AI companions” to children (ParentsTogether Action & Heat Initiative, 2025), proliferation of woke/anti-conservative ideology (Mazeika et al., 2025), and large-scale job displacement (Brynjolfsson et al., 2025). A transhumanist-led future could erode human dignity by pursuing a post-human agenda that rejects the inherent worth and nature of man. Avoiding these dangerous outcomes is essential for human flourishing and for earning the confidence of the public at home and beyond.

A PRINCIPLE-LED APPROACH

The America First approach should therefore advocate responsible leadership in AI technology, putting both America and the American people first. It is based on four principles:

Principle I: American Leadership. America must outpace the world in AI.

Principle II: Human Dignity. AI must put humanity and American values first.

Principle III: American Security. The United States must defend the homeland and the American people.

Principle IV: Government Efficiency and Readiness. The United States must use AI to streamline government processes, cut bureaucratic waste, and promote strategic AI readiness.

Principle I: American Leadership

America must outpace the world in AI. The AI boom has the potential to give Americans longer, healthier, more prosperous lives. The industrial buildout required by AI, if it is done well, can enable a more reliable electrical grid with more affordable power for all Americans. It can also create thousands more high-paying manufacturing jobs for American workers. To advance American leadership, government policies must:

1. Promote energy abundance. To continue training and deploying advanced AI systems, companies must exponentially scale compute (Sevilla & Roldán, 2024). This means not only building more AI data centers, but also the energy infrastructure needed to power them. If recommendations are not heeded, American AI companies may soon lack the energy capacity needed to stay ahead in AI, thus ceding our AI lead to China (OpenAI, 2025a). Bold action is therefore needed to ensure that America can develop and build energy infrastructure without raising costs on everyday consumers.

2. Ensure American AI hardware leadership and independence. Generations of failed leadership have ceded American dominance in manufacturing and cutting-edge technology. Today, America is completely dependent on Taiwan and other foreign countries for AI chip manufacturing, but it does not have to be this way (Flamm & Bonvillian, 2025).Restoring U.S. leadership in the manufacture of semiconductors, robotics, and advanced AI hardware can set the stage for a new industrial revolution and help create a golden future built in America.

3. Encourage widespread AI adoption. Having the best AI technology in the world is meaningless if it cannot be widely adopted throughout the economy. As identified in Executive Order 14320, Promoting the Export of the American AI Technology Stack (Exec. Order 14320, 2025), America must therefore not only invent the best AI technology, but also ensure its wide diffusion and adoption, domestically and globally, to create prosperity. The Biden Administration kneecapped diffusion and adoption of U.S. AI by attacking open-weight AI (AI models with freely available model weights), ceding global AI standards to foreign governments, and failing to support market-based solutions to reliability and robustness challenges that undermine trust in the technology (Exec. Order No. 14110, 2023). An America First AI agenda must chart a different path by reversing these failures.

4. Accelerate human-centered AI applications. Principled, responsible use of AI can help our kids learn skills to thrive in a fast-paced world, freeing them from woke curricula in dysfunctional public schools. As First Lady Melania Trump stated in the Presidential AI Announcement, we are poised to lead in A (The White House, 2025). Cutting government red tape can help AI drive breakthroughs in medicine. Americans can reap the benefits of AI without betraying our core values.

The following are actions policymakers could take to advance American AI leadership:

PROMOTE ENERGY ABUNDANCE.

• Streamline the construction of high-priority infrastructure projects. Create a new “Gold Permit” category that supersedes and replaces all federal permitting requirements for high-priority infrastructure projects.
• Reduce the impact of data center energy demand on consumers. Create incentives (e.g., fast-track permitting processes) for infrastructure developers to self-supply energy to power data centers.
• Streamline state-level permitting approvals. Repeal state-level laws that delay projects and impose regulatory overhead, including laws that adopt NEPA-like requirements.
• Implement demand-response programs. Accelerate grid interconnection for AI data centers and other large industrial energy consumers that voluntarily agree to participate in measurable demand-response programs.
• Reserve important supply chain components for emergencies. Create a strategic reserve – only to be used during emergencies – of high-voltage transformers, natural gas turbines, and other key components necessary for energy infrastructure supply chains.

ENSURE AMERICAN AI HARDWARE LEADERSHIP AND INDEPENDENCE.

• Promote domestic industries critical to AI hardware. Use existing funding allocations to strategically accelerate America's semiconductor fabrication, critical minerals refining, and advanced packaging ecosystems.
• Achieve semiconductor manufacturing independence. Create public-private partnerships to onshore semiconductor fabrication and advanced packaging capacity; encourage fabless American chip design companies (those that do not manufacture the chips they design) to use newly spun-up American foundries.
• Win the robotics race. Develop a national strategy for robotics that includes identifying regulatory and supply chain barriers to U.S. dominance in robotics manufacturing.
• Prevent China from indigenizing chip manufacturing. To promote the export of the American AI stack, implement a global ban on the export of key semiconductor manufacturing equipment and subsystems to China, including machines and equipment used for producing advanced logic and memory chips.

ENCOURAGE WIDESPREAD AI ADOPTION.

• Promote open-source AI. Leverage programs like the National Artificial Intelligence Research Resource to prioritize compute access for open-source AI model developers, enabling Little Tech developers to compete with Big Tech incumbents.
• Prevent China from setting global AI standards. Publish a report on the risks of a China-dominated AI ecosystem and share its findings with allies at multilateral fora.
• Ensure the U.S. leads in global AI standards. Generate enterprise integration standards to encourage the global use of American AI, rather than foreign AI.
• Solve reliability and robustness challenges in AI. Fund new research and development (R&D) on methods for ensuring AI systems used in high-security contexts reliably and robustly follow commander intent.

ACCELERATE HUMAN-CENTERED AI APPLICATIONS.

• Track the use of AI tools in schools. Study effective implementation of AI-assisted education in schools, identify barriers to adoption for AI-assisted education, and implement direct reporting on AI use in schools through existing survey mechanisms.
• Establish prototypes of responsible AI-assisted education. In partnership with local school districts, encourage universities to establish Responsible AI Tutor Prototyping Centers that promote thoughtful, responsible implementation of AI-assisted education in schools.
• Promote AI literacy in schools. Encourage integration of AI literacy in K-12 academic standards covering how to use AI, how AI works, and how to build AI systems and applications.
• Unlock government datasets for healthcare innovation. Open certain datasets, such as those within the National Institute of General Medical Sciences (NIGMS), to support AI research that drives medical innovation.

Principle II: Human Dignity

Innovation must put humanity and American values first.

AI requires unique design choices compared to previous technologies. Developers embed preferences directly into model behavior, thereby determining which values those systems propagate to billions of users. Big Tech companies get to decide what their chatbots optimize for, how they steer users, and what guardrails (if any) exist to prevent harmful outcomes.

Silicon Valley must not be allowed to poison our kids with slop and smut, flood the internet with woke propaganda, or replace American workers instead of making AI technology work for them. AI can transform American lives for the better, but we do not need to sell out our values to Big Tech to make it happen. To uphold human dignity, government policies must:

1. Protect children from harmful AI chatbot interactions. Nothing is more important than the sanctity of the American family. Although responsible use of AI in schools can enhance human dignity and children’s wellbeing, AI companies betray our values when they cut out parents and let their products show harmful content to our children (Knutsson, 2025). Empowering parents and creating strong guardrails against sexual material, pro-suicide outputs, and all other inappropriate or dangerous AI content is a top priority for the America First movement.

2. Promote transparency to ensure public accountability in AI. As AI becomes more prevalent in American life, AI behaviors and values matter more than ever. We have seen concerning signs. AI models have defamed conservatives like Senator Marsha Blackburn and Robby Starbuck (Blackburn, 2025; Flood, 2025). AI outputs have repeatedly reflected deep bias against conservatives (Rozado, 2023; Rozado, 2025). Despite commitments to the contrary, San Francisco AI executives have openly stated they will steer their models toward their preferred policy objectives (OpenAI, 2025b; Altman, 2025). Thankfully, the Trump administration has demonstrated leadership by issuing EO 14319, Preventing Woke AI in the Federal Government (Exec. Order 14319, 2025). However, more work must be done to ensure that everyday Americans have full visibility into how AI models are trained and what biases and objectives they contain.

3. Stand up for the American worker. America First policy is about believing in the American worker. AI will change how work is done, but that does not mean leaving hardworking people behind. Training and job programs can help Americans help themselves by learning new skills and using AI to automate paperwork and toil. As AI agents partner with American workers to automate repetitive tasks, companies should ensure that AI agents are reliable, trustworthy, and controllable. Just as we protect American workers from being undercut by illegal immigrant labor, we must protect them from being replaced by incompetent, unreliable AI.

The following are actions policymakers could take to uphold human dignity:

Protect children from harmful AI interactions.

• Implement app store and chatbot age verification. Provide guidance on technical methods for chatbot and app store providers to develop age verification systems and require age verification based on technical best practices.
• Report on dangerous interactions between chatbots and minors. Require companies to detect dangerous or prohibited conversations between minors and chatbots and report them to parents and the government.
• Prevent sexualized AI chatbot outputs to minors. Prohibit AI chatbots from engaging in conversations related to self-harm, suicide, and sexual content with minors through refusal training, age verification, and other methods.
• Prevent the output of child sexual abuse material (CSAM). Prohibit all AI-generated, sexually explicit images of minors.

Promote transparency to ensure public accountability in AI.

• Evaluate AI for false and deceptive outputs. Conduct evaluations of false, unreliable, or deceptive behavior of domestic and foreign AI, including woke ideology and CCP propaganda.
• Develop guidance on reliable, non-deceptive AI. Such guidance should provide a concrete evaluation baseline for developers to identify and correct deceptive behavior.
• Mandate disclosure of political AI compensation. Require AI companies to publicly disclose compensation received for services relating to how the system presents information about candidates, campaigns, or ballot measures.
• Require transparency on AI values and security. Require or incentivize AI companies to publish information about subjective judgments embedded into model behavior, as well as significant security incidents and system evaluations.

Stand up for the American worker.

• Track AI’s impact on skill demand. Develop and incorporate AI-specific metrics in surveys of job and labor market data and track the most in-demand skills in the AI era.
• Train Americans in high-demand skills. Expand programs in AI-impacted sectors with a focus on vocational training and registered apprenticeships.
• Promote reliability in AI agents. Develop and promulgate voluntary guidance on reliability, attributability, and trustworthiness of AI agents.

Principle III: American Security

The U.S. government must defend the homeland and the American people.

American companies face coordinated attacks from foreign governments. Chinese regulators impose arbitrary restrictions on U.S. firms while state-sponsored hackers target their intellectual property through distillation attacks, data poisoning, and outright theft of model weights (Wodecki, 2023; Sabin, 2024; National Security Agency, 2025). Individual companies cannot counter nation-state adversaries alone.

The American people are also vulnerable to malicious foreign uses of AI. AI profoundly affects the militaries and intelligence-gathering agencies of every country (Bondar, 2025). Countries that do not thoroughly and responsibly integrate AI into their national security will not just be “left behind,” they will be vulnerable to all manner of harm. Additionally, reckless or targeted gain-of-function research could become even more dangerous through AI-powered bioweapons programs (Kuiken, 2023). To bolster American security, government policies must:

1. Defend U.S. companies from adversarial foreign threats. American AI is the best in the world. Protecting our edge in AI means standing up to foreign adversaries who will try to punish U.S. companies for their success and steal American secrets for their own benefit. The Trump Administration can win deals that give U.S. companies a fair playing field abroad and help our AI companies achieve world-class security.

2. Track and resist hostile foreign AI. Foreign adversaries may try to use AI to harm the United States. Diplomacy and negotiation can show foreign countries the great rewards of fair treatment and friendship with America. The U.S. government should not tolerate foreign AI projects that threaten the American people. Our military and intelligence agencies can track foreign AI projects, assess them for dangerous actions or capabilities, and, if necessary, take action to neutralize the threat.

3. Develop resilience against AI-enabled foreign biothreats. AI has already demonstrated the ability to accelerate biochemical science. An AI team was even awarded the 2024 Nobel Prize in chemistry (Royal Swedish Academy of Sciences, 2024). Unfortunately, these capabilities could worsen biological threats. The Trump Administration has transformed America’s biodefense capabilities by going after reckless gain-of-function research and setting rules to keep gene synthesis technology away from dangerous criminals and foreign powers (Exec. Order No. 14292, 2025). Commonsense protections can further reduce threats to Americans and cement President Trump’s legacy as chief defender of American biosecurity.

The following are actions policymakers could take to bolster American security:

Defend U.S. companies from adversarial foreign threats.

• Ensure that American companies are treated fairly by foreign regulators. Consider retaliatory trade policies to deter foreign governments from imposing unfair trade barriers on U.S. AI companies.
• Develop best practices in AI security. Develop voluntary technical guidance, standards, and best practices for AI security that address unique threats such as distillation attacks, data poisoning, and model weight theft.
• Encourage adoption of AI security best practices. Adopt procurement policies that favor AI developers and compute providers who voluntarily adopt AI security best practices.
• Build the most secure AI data centers in the world for national security missions. Create a pilot program for a "maximum-security" data center for AI training and analysis with the most sensitive government data.

Track and resist hostile foreign AI.

• Track foreign, hostile uses of AI. Create an implementation plan for a crash program to detect, track, and assess AI projects in adversary nations.
• Prevent China from using AI against the American people. Through bilateral engagements with China on AI, ensure that China is not using dangerous or disruptive AI against the homeland.

Develop resilience against AI-enabled foreign threats.

• Track “sequences of concern” that can be used for gain-of-function research. Research and deploy cutting-edge techniques to screen nucleic acid sequences for potential pathogenicity, empower relevant agencies to implement gene synthesis screening requirements, and set security requirements for benchtop synthesis devices capable of printing DNA in low-resource settings.
• Prevent taxpayer dollars from funding gain-of-function research. Require federal research funding recipients to purchase biological materials from companies that adhere to established biological safety frameworks issued under President Trump’s executive orders on biosecurity.
• Evaluate America's resilience to biological attacks. Investigate various biological threat models, existing incident response plans and capabilities, and identify current gaps in American bio-resilience.

Principle IV: Government Efficiency and Readiness

AI must be deployed to streamline government efficiency and readiness.

The federal government is slow, old, and currently ill-suited to seize the potential of AI. That must change. The American people deserve a government capable of acquiring AI systems and using them to streamline public efficiency, automate administrative functions, and replace unneeded bureaucrats. The federal government must be staffed with the AI expertise needed to understand where the technology is headed and how it should be leveraged. To promote government efficiency and AI readiness, government policies must:

1. Automate government bureaucracy. Over decades, the administrative state has perpetually grown, leading to ever-increasing rules and regulations that burden taxpayers and undermine liberty. These problems are particularly pronounced in sectors like healthcare, where administrative overhead accounts for 25% of medical spending, and in education, where administrative staff have grown faster than student enrollment (Chernew & Mintz, 2021; Weinstein, 2023). AI offers a way out by automating and consolidating administrative processes. However, AI adoption in government cannot proceed effectively if the American people do not trust it; therefore, government must ensure that its use of AI does not harm Americans’ privacy or civil rights.

2. Bring top AI technology and talent into government. Although AI can be used to automate administrative processes, a labyrinth of internal red tape stops the government from hiring good people with the expertise needed to understand AI technology. The U.S. government should be using AI technology to defend our borders, stop cartels from trafficking fentanyl, and eliminate government waste. We cannot allow bureaucracy to get in the way of the government serving the American people. The U.S. government owes it to taxpayers to get the best technology and expertise to make government more efficient.

3. Build AI readiness in government. The federal government needs a nerve center of AI excellence that can monitor emerging capabilities, assess their implications for national security, and coordinate responses across agencies. Without this institutional capacity, agencies will react to AI developments piecemeal, missing threats and opportunities that require government-wide awareness.

The following are actions policymakers can take to promote government efficiency and readiness:

Automate government bureaucracy.

• Promote efficiency across the executive branch. Create an AI red-tape-cutting competition between federal departments, including leaderboards.
• Reduce administrative bloat in healthcare. Provide public and VA hospitals with an opportunity to embed AI engineers who can automate administrative functions without sacrificing healthcare quality.
• Reduce administrative bloat in education. Launch a pilot program on opportunities for AI to automate administrative functions in educational institutions.

Bring top AI technology and talent into government.

• Speed up government acquisition of advanced AI systems. Tools such as Other Transaction Authority and single-source contracting can be leveraged for efficient procurement of AI tools that can improve key government functions, including border security and military dominance.
• Expand hiring authorities for AI-related positions across government. Tools such as Direct Hire Authority and Excepted Service Hiring could be leveraged for top talent to enter and exit public service faster and with less red tape.

Build AI readiness in government.

• Authorize and fund the Center for AI Standards and Innovation (CAISI). CAISI is critical for ensuring that the federal government has strategic awareness of AI.
• Focus CAISI’s mission on innovation and national security. Ensure that CAISI’s mission adheres to the following: promoting and protecting U.S. AI leadership, measuring AI capabilities relevant for national security, and issuing voluntary guidance on AI security.

WAY FORWARD

The America First AI Agenda reflects AFPI's commitment to ensuring AI serves the American people. AI is a technology that offers great promise for improving the quality of life of Americans through healthcare, education, manufacturing, and American national security. Sound AI policy development requires that policymakers and the American people understand both what AI stands to offer, as well as how abuses of the technology could threaten human life and human dignity. An America First approach requires balancing innovation with security, economic growth with human dignity, and technological progress with American values.

The challenges facing AI policy vary significantly, from energy infrastructure bottlenecks to cybersecurity threats to protecting children online. AFPI will engage with experts across the industry and voters across the country to deliver policy outcomes that serve the American people. The policies outlined in this agenda provide a framework for American AI leadership. In the months ahead, AFPI will develop detailed implementation plans for each policy area, publish issue briefs that orient policymakers, and equip America First leaders with the analysis they need to govern AI for the benefit of the American people.