Top executives from OpenAI, Microsoft, AMD, and CoreWeave testified before the US Senate on May 8, urging action to maintain the nation’s lead in AI amid competition from China.

They highlighted the need for increased investment in AI infrastructure, education, and exports.

Microsoft President Brad Smith emphasized the importance of global AI adoption aligned with democratic values and raised concerns about Chinese AI advancements like DeepSeek.

OpenAI CEO Sam Altman stressed the need to expand data centers and power infrastructure to support AI growth.

AMD CEO Lisa Su and others called for policy changes to ease AI chip exports, promote domestic innovation, and enhance workforce training.

The hearing also addressed recent US export restrictions on AI chips to China.

🔗 Source: Reuters

🧠 Food for thought

1️⃣ Infrastructure gap represents a multi-trillion dollar competitive frontier

The Senate testimony highlights a critical but often underappreciated aspect of AI competition: physical infrastructure requirements are massive and growing exponentially.

McKinsey research indicates global demand for compute power will require approximately $6.7 trillion in investment by 2030, with $5.2 trillion specifically needed for AI-related data center infrastructure 1.

This explains why executives emphasized infrastructure as “critical” during testimony, as the scale of investment needed creates a strategic vulnerability where falling behind could be irreversible.

The competition extends beyond just building models to creating the physical capacity that enables AI development, with data centers potentially tripling by 2030 according to industry projections 1.

For context, this investment requirement exceeds what the US spent on all interstate highways, airports, and electrical grids combined over several decades, highlighting why executives are concerned about America’s ability to scale quickly enough.

2️⃣ Export control enforcement faces critical resource constraints

The debate around AI chip export restrictions reveals a major implementation gap that undermines American policy effectiveness regardless of approach.