Trump’s AI framework gets the big question right. Now comes the hard part.
Somebody in Washington finally asked the right question. Not “How do we regulate artificial intelligence?” but “How do we make sure America wins?” Those are not the same question. They never were. And the Trump administration deserves credit for knowing the difference.
The administration’s AI posture, examined recently by the Cato Institute, represents a deliberate repudiation of the Biden-era regulatory philosophy. Where Biden saw risk to be managed, Trump sees a race to be won. Where Biden built review boards and reporting requirements, Trump cleared the field and told American innovators to run. That is not recklessness. That is strategy.
Consider what the stakes actually are. Artificial intelligence is the defining technology of the coming generation. It will reshape military power, economic productivity, intelligence gathering, and national infrastructure. China understands this perfectly. Beijing is not pausing to convene a federal task force on algorithmic bias. It is spending, deploying, and winning ground while its rivals hold hearings. Every month of unnecessary regulatory delay is a gift to our adversaries delivered at American taxpayer expense.
“Every month of unnecessary regulatory delay is a gift to our adversaries.”
CLEARING THE RUNWAY
The administration moved early and moved decisively. Its January 2025 executive order instructed agencies to strip out policies that functioned as obstacles to AI development and to build a national strategy centered on American dominance in the field.1 That is the correct order of operations. You clear the runway before you fly the plane.
What the Biden administration never fully grasped was that precaution has a price. Every mandatory disclosure requirement, every pre-deployment review board, every new federal layer added cost and friction for American developers while imposing no comparable burden on foreign competitors. The result was a policy that protected no one and handicapped everyone building something worth protecting.
Republicans have long understood that the private sector, not Washington, is where transformative technology gets built. The administration’s preference for voluntary frameworks, industry cooperation, and public-private partnerships rather than a new federal bureaucracy reflects this understanding.2 Government should be a partner in this race. It should not be the referee who keeps stopping the clock.
WHERE CYBERSECURITY CHANGES THE CALCULUS
Here is where intellectual honesty is required. Even committed free-market conservatives must acknowledge that certain AI capabilities present genuine national security concerns that go beyond the ordinary risks of commercial technology. Frontier AI systems capable of identifying software vulnerabilities, accelerating cyberattacks, or assisting foreign adversaries in targeting critical infrastructure are not simply a faster version of the search engine. They represent a new category of risk.
The administration has recognized this. Recent executive action established a framework for voluntary government testing of the most advanced AI systems before public release, with federal agencies authorized to review frontier models for cybersecurity vulnerabilities.3 This is the right call, and it is the kind of targeted, security-focused engagement that conservatives should support: narrow, specific, tied to genuine threats rather than generalized anxiety about new technology.
The Cato analysts note, correctly, that voluntary frameworks carry their own risks. Yesterday’s voluntary review can become tomorrow’s de facto licensing system. That concern is legitimate and must be watched. But it is an argument for vigilance, not for refusing to address real national security threats. The solution to regulatory creep is institutional discipline, not institutional paralysis.
THE MARYLAND DIMENSION
For Marylanders, this debate is not abstract. The National Security Agency and U.S. Cyber Command sit at Fort Meade. The National Institute of Standards and Technology operates in Gaithersburg. Johns Hopkins Applied Physics Laboratory anchors some of the most sophisticated defense research in the country. Maryland is, by any reasonable measure, ground zero for the intersection of AI and national security.4
A pro-growth AI policy that keeps federal investment flowing, attracts private defense technology firms, and positions American innovation ahead of China’s is directly in Maryland’s economic and strategic interest. Maryland Republicans should be making this case loudly and often. The state has every reason to embrace the administration’s framework and every reason to resist the impulse of neighboring jurisdictions to pile on state-level restrictions that would complicate the national picture.
California and New York may have the luxury of treating AI regulation as an ideological signal. Maryland cannot afford that indulgence. The people who keep this country safe work here. The technology they depend on is being built right now, often with Maryland talent and Maryland resources. The question is whether state leaders will support that mission or create new headwinds against it.
THE CENTRAL QUESTION
The Cato analysis frames the administration’s fundamental challenge with precision: How do we prevent AI from being used against America without preventing America from leading AI?5 That is exactly right. Those two objectives can coexist. They do not have to be in conflict. But they require discipline, clear priorities, and a governing philosophy that does not treat innovation as a problem requiring management.
Remove the barriers. Beat China. Protect critical infrastructure. Use targeted cybersecurity reviews where genuinely warranted. Keep innovation in the private sector where it belongs. These are not contradictory goals. They are the only coherent strategy for a nation that intends to remain the world’s leading technological power.
The race is on. The administration, to its credit, knows it. Now the work begins.
ENDNOTES
1. Executive Order on Removing Barriers to American Leadership in Artificial Intelligence, White House, January 2025. https://www.whitehouse.gov/presidential-actions/2025/01/removing-barriers-to-american-leadership-in-artificial-intelligence/
2. Assessing the Trump Administration’s Executive Order on AI and Cybersecurity, Cato Institute. https://www.cato.org/blog/assessing-trump-administration-executive-order-AI-cybersecurity
3. National Security Agency and U.S. Cyber Command, Fort Meade, Maryland. https://www.nsa.gov/about/locations/fort-meade/
4. Cato Institute: Assessing Trump Administration Executive Order on AI Cybersecurity. https://www.cato.org/blog/assessing-trump-administration-executive-order-AI-cybersecurity
Andrew McConnell is a contributor for DirectLineNews, a project of the Montgomery County Republican Club
Direct Line News is a free-access publication committed to good governance, fiscal accountability, and the Republican reform tradition.
Direct Line News is a Maryland-based free-access publication committed to the Republican reform tradition.
About The Author