From Protecting America’s AI Advantage to Expanding It - Why Congress Must Learn to Play Offense
- William P Wilson
- Jul 3
- 9 min read
SIG Strategic Perspectives No. 1
A Strategic Policy Essay
By Semiconductor & Innovation Group LLC
Introduction
The ideas presented in this paper grew out of attending the Hudson Institute’s discussion, Securing America’s AI Advantage: A Discussion on U.S. Export Control Policy, featuring Senator Jim Banks (Indiana) and Chairman Brian Mast (Florida), moderated by Hudson Institute Senior Fellow Michael Sobolik.
Initially, the objective was straightforward: prepare a summary memorandum for clients. But as the discussion unfolded, a broader question emerged. Both speakers articulated a sophisticated strategy for protecting America’s technological leadership. They discussed export controls, advanced semiconductors, trusted technology partnerships, restrictions on remote access, insider threats, and the strategic importance of preserving the United States’ advantage in frontier AI compute. Collectively, they described an increasingly coherent national strategy for preventing America’s competitors—particularly the People’s Republic of China—from closing the technology gap. Yet listening carefully to this interesting dialogue raised a second more important question:
What should Congress do to ensure that America continues creating the next generation of technological breakthroughs?
The discussion devoted considerable attention to stopping harmful technology transfers. Comparatively little attention was devoted to accelerating American innovation itself. This paper is therefore not intended as a critique of the speakers. Quite the opposite. Their remarks provide the foundation for a larger conversation—one that asks what Congress ought to do after establishing a strong defensive strategy. Protecting today’s technological lead is indispensable. Sustaining tomorrow’s leadership will require an equally ambitious agenda focused on engineers, research, commercialization, manufacturing, and allied innovation.
The Imperative
America cannot simply prevent China from catching up. America must continue pulling ahead.
"Presently, more than 90% of the lawmakers 'energy' on capitol hill is focused on just defensive actions. . .reactions to things happening in technology and in geopolitics. A winning strategy should look more like 60-40. . .60 percent on offense, focused on winning the future."
Executive Summary
The Hudson Institute bicameral discussion demonstrated that congressional thinking on artificial intelligence has evolved significantly.
AI export controls are no longer viewed primarily as trade policy. They are increasingly understood as instruments of national security, industrial strategy, and geopolitical competition.
Throughout the discussion, Senator Banks and Chairman Mast returned repeatedly to several themes:
advanced AI compute has become a strategic national asset;
semiconductor leadership underpins economic and military power;
trusted allies should form a common technology ecosystem;
China must be denied access to frontier AI capabilities.
These arguments represent a mature thoughtful strategy for protecting America’s technological advantage.
Though it is wholly incomplete. The next challenge is equally important: building the innovation ecosystem that produces the next advantage.
I. The New Congressional Consensus
Perhaps the most striking feature of the discussion was how settled many of the foundational questions have become. Rather than debating whether export controls should exist, the conversation focused on how they should evolve. Senator Banks framed the issue in stark strategic terms.
“When you’re in a war, you don’t give your enemy your best weapons and technologies.”
He argued that advanced AI chips should increasingly be viewed in much the same way policymakers view advanced defense systems. Banks further explained that domestic demand should receive priority before exports to strategic competitors. “Before you ever sell chips to China or an adversarial company, you have to meet the market demand in the United States first… You can’t give better prices to Chinese customers than you give to American customers.” Banks was referencing his bipartisan GAIN AI Act (S. ) introduced in this Congress but a version also passed the Senate last year as a part of the 2026 NDAA. This is an important conceptual shift.
Congress is no longer treating frontier AI compute simply as a commercial product. It is increasingly treating compute as strategic infrastructure.
SIG Analysis
This represents one of the most significant changes in congressional thinking over the past several years. The policy debate is moving beyond export licensing toward broader questions of national resource allocation, industrial resilience, and strategic competitiveness.
II. NVIDIA and the Politics of Compute
No company appeared more frequently during the discussion than NVIDIA. That is not accidental. NVIDIA has become the embodiment of the broader policy tension confronting Washington. Chairman Mast praised NVIDIA as “a great American story.” He acknowledged Jensen Huang’s confidence that NVIDIA could successfully compete in China. Yet he drew an important distinction between corporate success and national success. “I believe Jensen Huang can go into China and win… Good for him. But that doesn’t mean the United States wins. China needs America’s compute.” This statement captures the central dilemma facing policymakers. The question is no longer whether American companies should succeed internationally. The question is whether success in strategic markets can unintentionally accelerate the capabilities of America’s principal geopolitical competitor. The age-old corporate rejoinder, first entering the DC lexicon during a 1953 U.S. Senate confirmation hearing for Charles Erwin Wilson, the president of General Motors who was nominated to be Secretary of Defense by President Dwight D. Eisenhower. When defending his substantial stake in GM that might compromise his judgement at Defense, Wilson argued strongly (though not explicitly) that "What's good for General Motors is good for America." It encapsulates a mid-20th-century belief that the success of massive private enterprise inherently translates to national prosperity. This is nearly identical to similar arguments made by technology and AI pioneer leaders today who directly equate their business success and profitability with "America's strategic interest." But, most of the business and technology leaders making these arguments today are too young to remember very problematic analogies to the mid-twentieth century when many major US manufacturing companies were directly involved in helping to build Germany's massive military expansion and rearmament.
Major US companies that significantly contributed to Germany’s 1930s military build-up provided automotive transport, strategic fuel additives, telecommunications equipment, and logistical computing technology.
The primary corporations involved operated through strategic partnerships, technology sharing, and large-scale ownership of German subsidiaries:
General Motors (GM): Through its German subsidiary, Adam Opel AG, GM manufactured military trucks and armored vehicles (such as the Opel Blitz) that served as the standard logistical backbone of the German army.
Ford Motor Company: Through its subsidiary Ford-Werke, Ford produced heavy trucks and military equipment for the German war effort.
Standard Oil of New Jersey: Provided the Third Reich with aviation technology and produced tetraethyl lead, a crucial gasoline additive. Captured records show German officials stated their aviation fuel and Blitzkrieg strategy were made possible by Standard Oil's experimental knowledge and production plants.
International Telephone and Telegraph (ITT): Held major stakes in German telecommunications and supplied the German military with switchboards, radar components, and artillery fuses.
IBM: Through its German subsidiary Dehomag (Deutsche Hollerith-Maschinen Gesellschaft), IBM provided tabulating equipment and punch-card technology. This technology was used to track census data and later helped to manage the logistical operations of the Nazi regime
From a strategic humility and communications perspective, it would be helpful for corporate leaders engaged in debates today about market access in China, to pay at least some deference to this history and include a reminder that US business has not always got it right. Elected leaders (especially US military veterans like Mast and Banks) are perhaps rightly worried that technologies developed in America might ultimately make Chinese weapons more effective, and more deadly. There is no denying the very chips at the center of our debate today, are the chips that WILL BE USED to refine targets and kill more Americans in the future. This is the lesson being digested by War Department planners right now, from the hot war in Ukraine. AI is the perfect tool for adjusting fire and constantly refreshing high value targets. Despite arguments about trade policy and the need for Chinese origin profits to fuel R&D efforts, technology business leaders should be cautious in dismissing these concerns from elected officials and defense thinkers. It is not an unreasonable opinion to worry that just a few dozen of the most powerful AI chips could mean many more casualties on future battlefields. It may be unavoidable, but they are not wrong to worry about this capability.
SIG Analysis
NVIDIA has become the case study through which Congress is redefining the relationship between commercial leadership and national security. Future debates over export controls, cloud computing, licensing, and trusted AI ecosystems are likely to continue using NVIDIA as the principal reference point.
III. Winning the AI Race
Chairman Mast repeatedly rejected the idea that the United States should seek technological parity with China.
“I’m an American. I want superiority to China—not parity.” He also challenged the idea that AI competition has a finish line. “The AI race is an indefinite race.”
Perhaps his most insightful observation concerned the value of even a small technological lead. “Whether we’re six days ahead or six months ahead, it matters.” In AI, time itself becomes a strategic advantage. Even modest leads in frontier capabilities may translate into significant advantages in cybersecurity, defense applications, scientific discovery, or military planning.
IV. Trust and Technology
One of the discussion’s most thoughtful exchanges concerned allied confidence in American technology.
Responding to questions about “kill switches” and technological dependence, Chairman Mast emphasized that America’s comparative advantage should be trust. “I want a handshake America… Your word is your bond.”
He argued that trusted partnerships require obligations on both sides. American companies must be transparent about what their technology does. Trusted partners must ensure that technology is neither diverted, reverse engineered, nor made available to adversaries. This is an important evolution in American technology diplomacy.
The objective is no longer simply exporting products. It is building trusted technology ecosystems.
Anyone tempted to reduce the technology debates on capitol hill to mere small-mindedness or reluctance to change should pay special attention to this section of the discussion. Chairman Mast, the Chairman of the House Foreign Affairs Committee is making a powerful point here about how America treats its partners and allies and the responsibility we have to earn their trust. These global partners are not just customers, they are co-laborers in a big project to power for the democratic alliance. Despite harsh criticism and what Senator Banks called "tough love" from the President of the United States, everyone understands the assignment. From Pax Silica (the US State Department's flagship economic security coalition spearheaded by Under Secretary for Economic Affairs Jacob Helberg) to the Senate's Multilateral Alignment of Technology Controls on Hardware (MATCH) Act (sponsored by Senator Jim Risch (R-Idaho), chairman of the Senate Foreign Relations Committee, with Senators Pete Ricketts (R-NE), and Andy Kim (D-NJ)) political leaders acknowledge allies need to move together and use their cooperation to win.
"The lesson of history is that collaboration is America's real superpower."
V. Congress Must Learn to Play Offense
This is where the discussion naturally leads. The speakers presented a compelling framework for protecting America’s technological leadership. The next challenge is expanding it. Throughout the event, Chairman Mast repeatedly returned to one simple idea. “You gotta be hungry.”
He illustrated the point personally, describing how his own children complete additional mathematics courses every summer—not because they are struggling, but because America cannot expect to compete globally without producing exceptional technical talent. Senator Banks similarly highlighted Purdue University and the importance of ensuring that America’s engineering programs continue producing the workforce required for semiconductor leadership. Those comments point toward a much larger policy agenda that has yet to be fully developed. Getting ideas and energy into this lane SHOULD be industry's highest priority.
Congress now speaks fluently about export controls.
It should speak just as fluently about innovation.
That means asking difficult questions:
How do we double the number of electrical engineers graduating each year?
How do we strengthen America’s research universities?
How do we expand graduate fellowships in strategically important fields?
How do we accelerate commercialization of federally funded research?
How do we deepen partnerships between universities, national laboratories, industry, and venture capital?
How do we ensure that semiconductor incentives are matched by comparable investments in people?
How do we add muscle to the White House's American AI Action Plan?
Export controls may slow China’s progress. Only innovation expands America’s lead.
VI. A Five-Pillar Offensive Strategy
Congress should complement its defensive agenda with an equally ambitious innovation strategy built around five pillars.
Talent. Expand engineering education, graduate STEM programs, apprenticeships, and technical workforce development.
Discovery. Increase investment in research universities, national laboratories, and mission-driven scientific research.
Commercialization. Improve technology transfer, startup formation, venture financing, and manufacturing scale-up.
Allied Innovation. Move beyond allied procurement toward joint research, standards development, and shared innovation ecosystems.
Strategic Demand. Use federal procurement, defense acquisition, and public-private partnerships to accelerate deployment of American technologies.
Taken together, these pillars represent an offensive strategy for sustaining long-term American leadership.
Conclusion
The Hudson Institute discussion demonstrated that Congress is rapidly developing a coherent strategy for protecting America’s AI advantage. That is both necessary and overdue. The next challenge is ensuring that Congress develops an equally coherent strategy for expanding that advantage. America’s greatest strategic strength has never been its ability to deny technology. It has been its unmatched ability to create it.
The semiconductor revolution, the internet, GPS, biotechnology, cloud computing, and artificial intelligence all emerged from an innovation ecosystem built on world-class universities, scientific research, entrepreneurial risk-taking, and competitive markets.
That ecosystem cannot be taken for granted.
Congress has made meaningful progress on the defensive half of the equation. The next chapter in American technology leadership is learning how to play offense.
Comments