Grok 
DeepAI 
During a rare on-the-record briefing with reporters in San Francisco, Altman expressed deep concern about the multifaceted nature of the US-China AI competition. He stated, according to CNBC, "I'm worried about China," and added, "There's inference capacity, where China probably can build faster. There's research, there's product; a lot of layers to the whole thing. I don't think it'll be as simple as: Is the U.S. or China ahead?"
Altman voiced skepticism about Washington's current strategy of tightening semiconductor export restrictions. When asked if reducing the number of GPUs reaching China would be reassuring, he was blunt, "My instinct is that doesn't work." He elaborated on his concerns, stating, "You can export-control one thing, but maybe not the right thing... maybe people build fabs or find other workarounds. I'd love an easy solution. But my instinct is: That's hard."
His comments come as the Trump administration continues to adjust its approach. In April, President Trump halted all advanced chip exports to China, but the US recently made exceptions for certain "China-safe" chips. This new policy allows companies like Nvidia and AMD to resume limited sales in exchange for giving the federal government 15% of their China chip revenue.
China's open-source strategy and its impact on OpenAI
Altman also acknowledged that China's advancements, especially through open-source models like DeepSeek, directly influenced a significant strategic shift at OpenAI. He revealed that this competition was a key factor in OpenAI's decision to release its own open-weight models - gpt-oss-120b and gpt-oss-20b - on August 5, the company's first such release since GPT-2 in 2019. Altman explained, "It was clear that if we didn't do it, the world was gonna be mostly built on Chinese open-source models. That was a factor in our decision, for sure. Wasn't the only one, but that loomed large."
The Chinese company DeepSeek made headlines earlier this year by developing its R1 model for just $5.6 million, a fraction of the cost of its Western rivals. This success is part of a broader trend, as Chinese open-source models now dominate global rankings. Moonshot AI's Kimi K2, MiniMax M1, Qwen 3, and various versions of DeepSeek R1 are now occupying the top spots among developers worldwide.
Strategic Implications of the warning
Altman's warning highlights a growing concern among experts that Washington's export-control strategy may be counterproductive. Rather than containing China's AI ambitions, these policies could inadvertently accelerate Beijing's push for technological self-reliance. Chinese companies are already actively developing domestic alternatives through suppliers like Huawei, and the government is encouraging the use of homegrown hardware by branding US-made chips as potentially unsafe.
The OpenAI CEO's candid assessment underscores the critical challenge for US policymakers: how to effectively compete with China in the AI race without implementing policies that ultimately backfire by spurring even more aggressive investment in China's domestic AI ecosystem.
