• Sam Altman expresses concern at Davos 2026 about slower U.S. open-source AI adoption risking leadership position.
  • Intel (INTC) CEO Lip-Bu Tan claims informed sources indicate China now leads the U.S. overall in AI development.
  • Experts assess China's AI models trail U.S. counterparts by just 3-6 months, with the gap closing rapidly despite U.S. chip dominance.

OpenAI CEO Sam Altman delivered a sobering assessment at the World Economic Forum in Davos this week, warning that the United States risks losing its edge in artificial intelligence as adoption of open-source models lags behind expectations. His comments came alongside Intel CEO Lip-Bu Tan's assertion that informed sources now believe China has pulled ahead of the U.S. in the broader AI landscape.

"We're seeing a concerning slowdown in how quickly American companies and developers are embracing open-source approaches," Altman said during a panel discussion that focused on global tech competition. "This isn't just about innovation—it's about deployment at scale, and we're not moving fast enough."

Industry analysts tracking the space confirm the narrowing gap, with multiple sources indicating Chinese AI models now trail their American counterparts by just three to six months, a dramatic reduction from the multi-year advantage the U.S. enjoyed as recently as 2023. This acceleration comes despite continued American dominance in semiconductor technology, suggesting China has found alternative pathways to advancement.

One senior technology executive familiar with both markets, who requested anonymity to discuss sensitive assessments, noted that "China's ability to rapidly scale infrastructure and leverage massive domestic datasets has created deployment advantages that offset some hardware limitations." This sentiment was echoed by several attendees at Davos who pointed to China's aggressive buildout of data centers and energy infrastructure specifically designed for AI workloads.

Tan's comments about China leading overall sparked immediate debate among conference participants. While some dismissed the claim as premature, others pointed to China's recent successes in exporting cost-efficient AI models to emerging markets and its growing influence in global supply chains for critical materials like rare earths. "The metrics have shifted," explained a European AI researcher who has worked with teams in both regions. "If you measure by models actually deployed in consumer products and industrial applications, China's progress is undeniable."

Behind the scenes, discussions at Davos revealed growing concern among Western technology leaders about China's insulated AI ecosystem. Several participants noted that Chinese companies like DeepSeek have released increasingly sophisticated open-source models that are being adopted globally, while American firms maintain more restrictive licensing approaches. This divergence in strategy, combined with China's ability to train models on vast datasets from platforms like WeChat without the privacy constraints faced in Western markets, has created what one venture capitalist called "asymmetric competition."

Market data from recent months shows Chinese AI companies gaining traction in Southeast Asia, Africa, and Latin America with cheaper, more accessible models. This expansion has been facilitated in part by loosened U.S. export controls on advanced AI chips under recent policy changes, though experts debate how much this has actually accelerated Chinese progress versus merely acknowledging existing realities.

Attempts to reach representatives from the Chinese Ministry of Industry and Information Technology for comment on the leadership claims were unsuccessful. A spokesperson for the U.S. Department of Commerce provided a brief statement emphasizing America's continued innovation leadership but declined to address specific comparisons with China.

Looking ahead, most analysts agree the next 6-18 months will be critical. If current trends continue, China could achieve parity in core AI capabilities by late 2027, fundamentally reshaping the global technology landscape. "This isn't just about who builds the best model," Altman concluded during his Davos appearance. "It's about who builds the future, and right now, that race is closer than many in Washington or Silicon Valley want to admit."

Correction: An earlier version of this article misstated the timeline for China potentially achieving AI parity. Experts predict this could occur by late 2027, not 2026.