America AI Exports Program Comments

Docket No. ITA-ITA-2025-0070

Section A — Respondent Background

1. Identify and describe who you represent and explain why you are providing input to this RFI.

The Wahba Initiative for Strategic Competition (WISC) is an applied policy and research initiative based at New York University’s Development Research Institute. WISC focuses on strategic competition, geoeconomic statecraft, technology governance, and the global deployment of trusted infrastructure. Its work includes policy research, public-private convenings, and engagement with governments, industry, and international partners.

WISC is not a commercial exporter of AI technologies. Rather, it operates at the system level—analyzing how technology architecture, financing, regulation, and geopolitics interact to shape global adoption. We submit this response to support the Department of Commerce in designing an American AI Exports Program that is commercially viable, strategically effective, and trusted internationally.

Our perspective draws on experience across government, industry, academia, and international engagement, and is intended to help ensure that the Program succeeds in real-world deployment rather than remaining aspirational.

2. If you represent a potential American AI exporter, what goods or services does your organization offer?

Not applicable.

3. If you represent a potential American AI exporter, how could this Program support your exports?

Not applicable.

Section B — The AI Technology Stack

4. Should the components of the AI-technology stack described in E.O. 14320 be clarified or expanded upon?

Yes. The stack should be clarified to reflect how AI is actually experienced by countries—as an integrated system that begins with connectivity and power and culminates in models and applications.

Commerce should explicitly clarify that “networking” includes the full telecommunications and interconnect substrate required for AI at scale, including fiber backbones, subsea cable connectivity and landing infrastructure, data center interconnects, routing, spectrum access, and core network security. These elements are essential for data replication, workload distribution, resilience, and sovereign control.

The stack should also be explicitly energy-aware, addressing credible power availability, resilience planning, and lifecycle operating costs. Without trusted connectivity and reliable energy, AI systems cannot be secure, scalable, or sustainable in priority foreign markets, regardless of the quality of the models or compute deployed.

Finally, the stack should be clarified to allow qualified components from allied and partner countries where those components meet U.S.-defined standards for cybersecurity, interoperability, transparency, and compliance. In practice, many successful deployments will combine U.S.-origin compute, cloud services, and models with allied contributions in telecommunications, energy infrastructure, systems integration, cybersecurity operations, and sector-specific applications. Allowing such integration increases deployability, reduces cost, and accelerates adoption while maintaining security through standards-based controls.

5. What factors should guide the evaluation of each component of the tech stack?

Pass.

6. What challenges might a consortium face in developing a full-stack proposal?

Pass.

Section C — Consortia Membership and Formation

7. What guidance should be provided on how consortia are formed and governed?

Guidance should be principles-based and industry-neutral, emphasizing: (i) modularity and interoperability; (ii) clear governance and accountability; (iii) security-by-design; (iv) transparency and auditability; and (v) demonstrated execution capability in foreign markets. Commerce should avoid guidance that implicitly crowns a “natural” leader (chip, cloud, model, or telecom) and instead define the outcomes consortia must deliver: deployable, trusted, sovereignty-respecting systems.

8. On consortia membership and composition:

a. Eligibility of consortium members

Eligibility should be based on the ability to meet requirements for cybersecurity, interoperability, transparency and auditability, and compliance readiness—rather than sector or size. Preserving the dynamism of the U.S. ecosystem requires enabling participation across segments, including integrators and security firms, not only infrastructure providers.

b. Eligibility of a consortium as a whole

A consortium should demonstrate: (i) the ability to deliver a complete package through direct provision or contractual integration; (ii) credible deployment capacity and lifecycle support; (iii) a clear governance structure with accountable leadership; and (iv) a compliance plan that makes export controls and end-use requirements operationally realistic.

c. Modularity in consortium formation

Yes. Commerce should encourage modularity by requiring proposals to describe substitutability at key layers—compute, cloud services, connectivity/security, and model governance—and to show how partner countries can adopt components without lock-in. Modularity reduces gatekeeper risk and increases exportability across diverse markets.

d. Changes in consortium membership

Pass.

9. On the role of foreign companies and countries:

a. Participation of foreign entities

Allied and partner-country firms should be permitted to participate when they meet clear standards for cybersecurity, interoperability, auditability, and compliance with U.S. export control and end-use requirements. A standards-based approach—rather than nationality-based exclusion—expands reach, strengthens coalition alignment, and prevents default to adversarial turnkey systems.

Foreign entities should be permitted where necessary for localization, integration, operations, or sector delivery and where they meet standards for cyber protection, transparency, auditability, and compliance. Excluding qualified local entities can slow adoption and reduce installations—the decisive metric of long-term influence.

b. How foreign entities should be involved

As integrated delivery partners selected through transparent consortium governance, with clear contractual provisions covering cybersecurity controls, audit rights, and compliance obligations.

c. Role of foreign governments

Foreign governments should play a defined but limited role in consortium development, focused on sovereign requirements and enabling conditions rather than commercial selection.

Host-country governments should articulate national priorities related to data governance, auditability, critical infrastructure protection, and regulatory compliance so that deployments strengthen national agency rather than create new dependencies.

Where a country’s firms or infrastructure providers are incorporated as components of a particular AI export package, the host government should also coordinate its development finance, export credit, and development assistance tools—including DFIs, export credit agencies, and relevant aid programs—to support deployment, local integration, and long-term sustainability.

Such coordination can accelerate installations, reduce execution risk, and reinforce alignment, while preserving the principle that consortia remain industry-led and commercially governed.

d. “Trusted partner” framework

Yes. A trusted-partner framework could reduce friction, speed deployment, and strengthen confidence. Criteria should include cyber maturity, procurement integrity, export-control cooperation, IP protection, and enforceable governance standards, including auditability. Benefits could include faster licensing pathways, prioritized financing, standardized technical assistance, and streamlined compliance processes.

10. Ensuring consortia are industry-led:

a. Designation of a lead entity

Yes. A lead entity should be accountable for execution and compliance and capable of coordinating multi-layer delivery. Suitability should be determined by governance capability, execution track record, and operational resilience—not by industry segment.

b. Federal role in consortium formation

Government should set principles, evaluation criteria, and guardrails, while allowing industry to form consortia. The federal role should focus on de-risking, convening where helpful, and enabling financing and regulatory clarity—without picking winners.

Section D — Foreign Markets

11. Priority countries or regions

Priority markets should be evaluated through a strategic portfolio lens rather than a single hierarchy. High-priority candidates include:

• Key security partners where digital infrastructure underpins alliance interoperability, deterrence, and crisis response.
  • Friendshoring and economic-resilience hubs central to supply-chain diversification and industrial capacity.
  • Energy-abundant and infrastructure-ready partners capable of supporting AI at scale through reliable, affordable power.

Markets where these factors converge—security relevance, supply-chain importance, and energy availability—should be viewed as especially attractive early candidates for full-stack deployment.

At the same time, the Program should recognize the developmental importance of broad inclusion. A U.S.-led AI tech stack that is modular, standards-based, and sovereignty-respecting can deliver substantial development benefits by expanding access to digital infrastructure, improving productivity, and strengthening institutional capacity. Every effort should therefore be made to extend access to such systems to as many countries as standards and security considerations allow. Where the United States does not offer a viable, deployable alternative, countries will not wait; they will adopt authoritarian-sponsored stacks that embed long-term dependencies and governance risks. Broad, responsible deployment of a trusted U.S.-led stack is thus both a development imperative and a strategic one.

12. Tradeoffs between country and regional prioritization

Pass.

13. Criteria for evaluating priority markets

Key criteria include energy availability and reliability, telecom integrity and routing resilience, regulatory capacity (data governance, cybersecurity, procurement), ability to sustain operations financially, and the presence of strategic competitive pressure.

Section E — Business and Operational Models

14. Business, operational, or ownership models to prioritize

The government should prioritize models that are sustainable, sovereignty-respecting, and resilient, with clear governance over data, auditability, and security controls. The Program should favor models that avoid lock-in and support local capacity building while maintaining safeguards for controlled technologies. Commerce can encourage such models by making them explicit evaluation criteria and aligning federal support with lifecycle sustainability rather than initial installation alone.

15–16. Documentation and non-traditional partners

Pass.

Section F — Federal Support

17. Most useful federal support mechanisms

The most consequential supports are those that determine whether installations occur:

• Fast, predictable financing delivered on competitive timelines.
  • Menu-based tools rather than bespoke processes.
  • Technical assistance addressing telecom readiness, energy planning, cybersecurity, governance, and workforce capacity.
  • Regulatory and licensing clarity, including predictable export-control pathways.

Economic Diplomacy Action Group coordination is essential because winning installations requires aligned diplomacy, financing, and execution.

a. Additional support mechanisms

Yes. Structured capacity-building for host-country governance and cybersecurity, and deployment support for permitting, procurement, and operations.

b. Operational changes needed

Yes. Tools must operate faster, more predictably, and in a coordinated manner. The Program will fail if U.S. processes cannot match the speed and simplicity of subsidized turnkey competitors.

18. Additional requirements for access to support

Access should scale with commitments to security-by-design, transparency and auditability, modularity and interoperability, and credible lifecycle support. Requirements should be clear and predictable.

Section G — National Security Regulations

19. Ensuring compliance with national security regulations

Compliance should be embedded through architecture-first safeguards—trusted compute, enforceable governance controls, strong cybersecurity, and auditable procedures. Partner-country legal capacity and diversion risk should inform what can responsibly be provided.

20. Reducing dependence on technologies of concern

By making the U.S. offer deployable and attractive: modular, sovereignty-respecting, secure, and financially competitive. The objective is expanding U.S.-aligned installations so future AI adoption rests on trusted foundations rather than captive channels.

21. Additional national security considerations

Avoiding lock-in to opaque systems is itself a security benefit. The Program should prioritize architectures that protect data governance, auditability, and routing sovereignty, while sustaining U.S. competitiveness across all segments of the ecosystem.

Section H — Evaluating Proposals

22. Evaluation factors

Commerce should assess strategic impact, feasibility (telecom and energy readiness), modularity and interoperability, security-by-design, auditability, speed-to-deploy, and scalability. The strongest proposals are those that can win in real markets and remain sustainable.

23. Non-consortium contributors

Pass.

24. Tradeoffs in selecting more or fewer consortia

Selecting too few consortia risks concentration and reduced adaptability. Selecting more consortia can increase competition and resilience, provided security and execution standards remain clear and enforceable.

25. Additional competitiveness factors

Competitiveness depends on installation success. Commerce should prioritize speed, simplicity, financing, and operational viability in diverse environments, while preserving U.S. innovation dynamism across segments.

Section I — Additional Information

26. Participation by non-consortium companies

Pass.

27. Promoting high-quality technical standards

Standards should be advanced through deployable architectures—interoperable, auditable, and secure-by-design systems embedded in procurement, financing, and technical assistance.

28. Additional considerations

The Program’s success will be determined by whether it wins installations. That requires synchronized financing, credible energy and telecom planning, and sustained operational support. Without speed and deployability, even superior technology will lose to subsidized turnkey alternatives.