The United States has built many of the world’s most advanced AI models and the chips that power them — yet it may still lose the global AI race. That surprising warning comes from economist Paul Prentice, a senior fellow at the Independence Institute, writing in American Thinker. His argument: America’s historic edge has come not just from inventing breakthrough technologies, but from spreading them across the economy. And today, that diffusion is where the U.S. is falling short.
Closed models aren’t enough
Prentice notes that the U.S. has developed many of the world’s most capable closed-source AI systems — like those behind ChatGPT and Claude — and designed advanced chips that supply extraordinary computing power. Yet that strength hasn’t translated into an equally competitive open-source ecosystem that businesses can use to build their own products. Open-source AI models, which are freely downloadable and adaptable, allow startups, manufacturers, hospitals, and farmers to customize AI for their specific needs without paying a provider for every query or sending sensitive data to a third party.
That adaptability, Prentice argues, is what turns a technological breakthrough into economy-wide growth. He points to historical parallels: electricity transformed productivity when factories reorganized around it, and the internet created far more value when businesses used it to launch new services. AI, he says, is the latest test of that formula — and America isn’t passing it.
China’s open models are catching up fast
The economic case for open-source AI would be compelling on its own, Prentice writes, but China makes it urgent. Chinese developers are offering capable, inexpensive models that businesses can adapt to their needs, and they’re competing for global market share. He cites Moonshot AI’s Kimi K3 as a striking example. In one broad composite evaluation, Kimi ranked ahead of Anthropic’s Claude Opus 4.8 — and it costs 40 percent less to run. While it doesn’t outperform the best American systems on every task, its rise undermines the assumption that superior closed models alone will preserve America’s lead.
Stanford researchers have also documented growing adoption of Chinese open-weight models, including by American companies, Prentice reports. As more businesses build on those models, developers create compatible tools, workers acquire model-specific skills, and investors fund complementary products. Those network effects, he warns, could create a durable advantage that’s hard to dislodge.
Regulation has costs
Prentice addresses concerns about misuse of open models, but he stresses that regulation itself carries economic costs. He cites Oxford professor Carl Benedikt Frey in the Wall Street Journal, who argues that China’s rapid AI progress stems from its ‘move fast and regulate later’ approach — allowing firms to scale in regulatory gray zones before intervening later. That, Prentice says, is a very different dynamic from Washington’s current debate over whether to crack down on AI software.
Broad restrictions on American open-model development would raise barriers to entry, protect incumbent providers, and steer cost-conscious businesses toward foreign alternatives that remain available, he argues. Safeguards should instead focus on specific, credible threats — while preserving competition and experimentation across the much larger market for beneficial uses.
Industry is already responding
Prentice points to the Open Secure AI Alliance as evidence that the market is already developing ways to manage risks. NVIDIA, Microsoft, IBM, Cisco, Palantir, Hugging Face, and others are collaborating on open tools that make AI systems easier to test, adapt, and secure. Shared tools can lower security and compliance costs across the market, he says, and the alliance shows that companies with different business models can develop common safeguards without treating openness as a defect.
His advice to policymakers: preserve that competitive environment. Instead of broad restrictions, target demonstrated risks and let private investment and industry coordination expand competitive American alternatives. The U.S. captured the full potential of electricity and the internet only by pairing invention with diffusion, he concludes — and it will capture AI’s economic potential only if it does the same again.
Source: www.americanthinker.com — https://www.americanthinker.com/blog/2026/08/america-s-ai-lead-depends-on-unleashing-innovation/