The US is losing its lead in AI

Summary of The US is losing its lead in AI

by The Verge

34m•July 20, 2026
Listen to Episode

Overview of The US is losing its lead in AI

This Vergecast episode examines the intensifying AI race between the U.S. and China, with Hayden Field and Lauren Feiner joining David Pierce to unpack how quickly Chinese AI labs are catching up to U.S. frontier models. The discussion centers on the technical and political stakes of that shift: distillation, export controls, model access, and what it would mean for U.S. power if China’s models became equal to or better than those from OpenAI, Anthropic, Google, and other American labs.

Key Points

China is no longer viewed as far behind

  • The old assumption that Chinese AI is “six months behind” is increasingly outdated.
  • According to Hayden Field, six months is now more like a worst-case lag scenario, not a reliable baseline.
  • Some everyday-use models may already be on par with U.S. offerings, or close to it.

Distillation is a major flashpoint

  • Distillation = training a smaller/newer model by learning from a larger, more capable one.
  • It’s a fast, cost-effective way for newer labs to catch up.
  • U.S. companies are worried Chinese firms have used distillation to accelerate progress by querying frontier models at scale and using the outputs to train their own systems.
  • Anthropic, OpenAI, and others see this as one of the main threats to their lead.

DeepSeek changed the conversation

  • DeepSeek’s emergence showed that highly competitive models could be built more efficiently and cheaply than expected.
  • That undermined the theory that restricting access to advanced U.S. chips would significantly slow China’s AI progress.
  • The episode suggests that compute restrictions alone may not be enough to preserve U.S. advantage.

Washington treats AI competition as a permanent geopolitical issue

  • Lauren Feiner notes that U.S.-China tech competition is a recurring priority in Washington.
  • The political framing has shifted over time, but the core concern remains: keep China from overtaking the U.S. in strategic technologies.
  • Export controls, chip policy, and AI regulation are all part of that broader contest.

The stakes go beyond product quality

  • The guests emphasize that the issue is not just “who has the best chatbot.”
  • AI is expected to be embedded into critical systems:
    • telecom
    • health care
    • defense
    • cybersecurity
    • surveillance
  • If Chinese models become dominant, the concern is that U.S. companies and even allies may depend on foreign AI infrastructure.

What Happens If China Wins?

The core fear in Silicon Valley

  • If a Chinese model becomes clearly superior, American AI firms could lose market share, influence, and strategic leverage.
  • U.S. companies worry about being forced to build on “second-best” models, which could affect competitiveness across the economy.
  • There’s also concern about data privacy, surveillance, and geopolitical influence.

The analogy to TikTok and Huawei

  • Lauren compares the AI debate to past U.S. anxieties over TikTok and Huawei:
    • Chinese tech dominance can become a national security issue.
    • Policymakers may feel pressure to block or restrict adoption even when the product is technically superior.
  • The difference with AI is that the consequences would be far more deeply integrated into infrastructure and everyday life.

How This Strengthens the U.S. AI Labs

  • The episode argues that the China competition gives U.S. frontier labs more leverage with policymakers.
  • Their message: let us move fast, avoid heavy regulation, and we’ll keep the U.S. ahead.
  • At the same time, those same companies are often asking for regulation in other contexts to limit liability and safety risks.
  • That creates a tension: they want regulation when it protects them, but not when it slows them down.

Political Outlook

Midterms could make the issue more visible

  • As the U.S. heads toward midterm elections, AI could become a stronger political talking point.
  • Congress is already under pressure over:
    • AI data centers
    • labor and energy concerns
    • safety and accountability
  • A visible “China is catching up” moment could be used to argue against tighter regulation and in favor of letting labs accelerate.

The Trump administration’s relationship with AI is fluid

  • The hosts describe the administration’s stance as inconsistent and rapidly shifting.
  • Industry-government alliances may hold for now, but could change quickly if one side feels crossed.
  • Much of the policy push may end up happening through executive action rather than legislation.

Takeaways

  • China’s AI progress is now a real, immediate strategic concern, not a distant possibility.
  • Distillation is one of the biggest technical and policy worries because it helps fast followers close the gap cheaply.
  • Chip export controls alone may not be enough to slow China’s progress.
  • The U.S. AI industry is using the China threat to argue for more freedom and fewer restrictions.
  • The political fight over AI is likely to intensify as the midterms approach.

Other Headlines Mentioned in the Episode

  • The Odyssey had a massive opening weekend, grossing $264.1 million worldwide.
  • The European Commission issued a record fine against AliExpress over unsafe and counterfeit goods.
  • Jensen Huang’s leather jacket sold at auction for $960,000.

Bottom Line

The episode frames the U.S.-China AI race as a high-stakes contest over technology, power, and policy. The biggest takeaway is that China may already be close enough that the old assumptions about American dominance no longer hold—and that uncertainty is reshaping both Washington’s politics and Silicon Valley’s strategy.