O'Leary says China fear will unite North America on AI, energy
Kevin O'Leary asserts that competition with China in AI and energy will unite North America economically. As Chinese AI models capture 58% of token usage among U.S. firms, industry leaders like Jensen Huang argue for innovation over bans. The data shows a strong market reliance on Chinese technology despite political tensions.

*this image is generated using AI for illustrative purposes only.
Investor Kevin O’Leary stated on July 26, 2026, that growing competition with China in artificial intelligence (AI), energy, and critical infrastructure is likely to push the United States, Canada, and Mexico toward deeper economic cooperation, despite existing political differences. O’Leary argued that the accelerating race for technological dominance and resource security makes regional integration an economic necessity rather than just a policy preference.
In a post on X, O’Leary wrote, "I believe fear of China, not politics, will ultimately bring North America back together." He emphasized that economic reality would outweigh political disagreements because the three nations are stronger when they collaborate. He highlighted the distinct comparative advantages of each country: Canada provides essential energy resources and critical minerals, the United States contributes technological innovation and economic scale, and Mexico remains a key component of North America’s manufacturing and economic network.
Market Dynamics and AI Competition
The call for unity comes amid intensifying competition in the AI sector. Data from OpenRouter indicates that Chinese AI models surpassed U.S. rivals in token usage for the first time, capturing 58% of usage among U.S. firms. This shift was driven by the growing adoption of models such as DeepSeek, underscoring China’s rising presence in the global AI market.
| Metric | Value | Context |
|---|---|---|
| Chinese AI Token Share | 58% | Usage among U.S. firms on OpenRouter |
| Key Driver | DeepSeek adoption | Rising presence of Chinese models |
Industry Perspectives on Regulation
Industry leaders have offered differing views on how to address this competition. Jensen Huang, CEO of Nvidia Corp., and Anthony Pompliano, CEO of Professional Capital Management, both advocated for competition through innovation rather than government restrictions. Huang praised Chinese AI models as "excellent," arguing that open-source technology could expand AI adoption while boosting demand for Nvidia’s chips and computing infrastructure. He rejected concerns about security risks, stating there is "zero possibility" China would run U.S. companies off the road, noting that models could be customized and isolated.
Pompliano similarly opposed potential bans on Chinese AI models, urging U.S. companies to build superior technology instead of relying on intervention. "Stop complaining and start competing. People won’t use Chinese models if American companies build something better," he said.
What the Numbers Show
The data reveals a significant divergence between political rhetoric and market behavior. While political leaders may emphasize restrictions, the 58% token share held by Chinese models among U.S. firms suggests that cost-efficiency or performance advantages are driving adoption. This trend highlights the challenge of decoupling supply chains in the AI sector, where access to advanced models and underlying compute infrastructure remains a critical competitive factor for North American firms.
How might the 58% adoption rate of Chinese AI models among U.S. firms influence upcoming U.S. legislative proposals regarding data sovereignty and tech decoupling?
What specific regulatory frameworks could the USMCA nations establish to harmonize AI safety standards while maintaining competitive advantages against China?
How will Nvidia's strategy of promoting open-source adoption impact its supply chain resilience if geopolitical tensions lead to stricter export controls on Chinese entities?

























