San Francisco – Sam Altman, the Chief Executive Officer of OpenAI, has issued a sharp warning that the United States risks falling behind in the rapidly escalating artificial intelligence (AI) race if it continues to underestimate China’s advancements in the field. Speaking at a media interaction in San Francisco, Altman cautioned that U.S. reliance on export restrictions as a defensive measure will not provide a reliable or sustainable solution.

“I Am Worried About China”

In his candid remarks, Altman said, “I am worried about China,” underscoring that the competition between Washington and Beijing in AI development is far more complex than many policymakers in the U.S. realize.

He explained that while American policymakers and tech leaders are often focused on export control regimes and tightening semiconductor supply chains, China has been steadily accelerating its research, product development, and commercial deployment of AI technologies. This has made it increasingly difficult to judge which nation currently has the upper hand.

Export Controls “Not a Winning Strategy”

The U.S. has long viewed advanced semiconductors as the backbone of cutting-edge AI systems. In recent years, successive administrations – first President Joe Biden, and now President Donald Trump – have moved aggressively to limit China’s access to these chips. Biden began by tightening licensing rules, while Trump has gone further by halting entire categories of semiconductor exports to Chinese firms.

But Altman expressed skepticism that these policies will yield the intended results. “My instinct says this approach will not work,” he said. “You can try to control the export of one thing, but maybe that’s not the right thing to control. People may find alternative solutions.”

He emphasized that while restrictions may create short-term obstacles, they cannot permanently block innovation in a country as vast and resourceful as China. “I wish there were an easy solution,” he added, “but my intuition says this is going to be extremely difficult.”

China’s Rising AI Ecosystem

Altman pointed out that Chinese companies and research institutions are investing heavily in AI infrastructure, not only building domestic expertise but also accelerating the rollout of new consumer products, enterprise tools, and advanced research models.

In particular, China’s strength in open-source ecosystems has given it a strategic edge. Many Chinese labs are making their AI models freely available, encouraging global developers to adopt and improve upon them. This openness, Altman admitted, has forced OpenAI to rethink its own strategy.

OpenAI’s Strategic Pivot

Until recently, OpenAI maintained a reputation for secrecy, choosing not to release its core models or share the underlying weights that power them. This was motivated by both safety concerns and the desire to preserve a competitive advantage.

But China’s growing influence, particularly through the availability of open-source systems capable of generating deepfakes and other advanced applications, changed the calculus.

Altman revealed that OpenAI realized it could not afford to remain closed while competitors gained traction worldwide. “It became clear that if we didn’t release open-weight models, then countries around the world would simply turn to Chinese open-source models,” he explained.

As a result, OpenAI launched its first two open-weight language models earlier this month – a landmark shift in company policy. These text-only models are designed to help developers, researchers, and businesses build products at lower cost, marking a significant move toward democratizing access to AI.

A Competitive, Uncertain Future

The decision underscores the high-stakes environment in which AI development is unfolding. For years, American companies like OpenAI, Google DeepMind, and Anthropic have dominated global discourse, setting benchmarks with their large language models. But China’s rapid scaling – from state-backed investment to private sector innovation hubs in Shenzhen, Beijing, and Shanghai – is beginning to challenge that supremacy.

Altman’s comments suggest a recognition that the AI race is not zero-sum but rather a multi-front competition where secrecy, regulation, and protectionism may prove inadequate. Instead, innovation, collaboration, and transparency will play critical roles in shaping global leadership.

The Broader Geopolitical Stakes

The stakes extend far beyond technology. AI is increasingly seen as a pillar of national security and economic power, influencing industries from defense to healthcare to finance. As Washington and Beijing jostle for dominance, both nations are aware that leadership in AI could determine the balance of power in the 21st century.

For the U.S., Altman’s warning is a call for policymakers to move beyond defensive trade measures and instead invest more heavily in domestic research, AI education, talent retention, and infrastructure. For China, the remarks validate the effectiveness of its aggressive state-led approach to funding research and encouraging public-private partnerships.

Conclusion

Sam Altman’s stark message is that the AI race between the U.S. and China is not a battle that can be won through export bans alone. While American leaders continue to restrict chip shipments, China is building momentum through research, open-source strategies, and product innovation.

OpenAI’s own pivot to open-weight models illustrates how seriously the company is taking the Chinese challenge. It reflects a new reality: in a world where global developers have choices, openness may be the only way to remain relevant.