The White House has renewed its push to maintain the US lead in AI as competition from China intensifies.

Michael Kratsios, Director of the White House Office of Science and Technology Policy, told a Senate subcommittee that the US lead in AI over China has narrowed significantly since 2020.

He said the US risks losing its dominance in AI and emphasized the need to export integrated “AI tech stacks”—comprising chips, algorithms, and applications—to allied countries.

Kratsios added that exporting these full AI stacks is seen as key to limiting China’s global market share in AI technology.

He noted that his office is working with the Commerce Department to provide further details on this strategy as part of the administration’s AI Action Plan.

The plan also calls for the defense and commerce departments to work with allies on export controls and to prevent adversaries from accessing US defense supply chains.

.source-ref{font-size:0.85em;color:#666;display:block;margin-top:1em;}

🔗 Source: South China Morning Post

🧠 Food for thought

Implications, context, and why it matters.

Investment patterns from 2017 predicted today’s closing AI gap

The White House’s acknowledgment that China has closed the AI gap “significantly” since 2020 reflects investment shifts that began years earlier. By 2017, China already accounted for 48% of global AI venture capital, surpassing the US’s 38% share, signaling a shift in global AI development momentum1. China’s government launched its New Generation Artificial Intelligence Development Plan in 2017, targeting a $150 billion AI industry by 2030 with massive state backing2. This early surge in Chinese investment followed what observers called a “Sputnik moment” when AlphaGo defeated a top Go player, drawing over 280 million viewers in China and spurring national AI enthusiasm3. The current US focus on exporting “cohesive AI tech stacks” represents a response to competition dynamics that were already evident half a decade ago.

China’s data advantages create structural competition challenges

The emphasis on exporting integrated AI tech stacks reflects recognition that China holds advantages in AI development that isolated export controls cannot address. China benefits from a population of 1.38 billion and extensive surveillance infrastructure with 176 million cameras, providing vast data for AI training that the US cannot easily replicate1. While the US leads in advanced supercomputers (5 out of 10 globally), China operates more total supercomputers (227 vs. 109), demonstrating a quantity-versus-quality dynamic in computational resources1. Previous export control efforts have shown mixed results, with leading US semiconductor equipment firms actually seeing revenue growth in China even after controls were implemented4. The tech stack approach acknowledges that effective AI competition requires integrated ecosystems of hardware, algorithms, and applications rather than controlling individual components. ……