Eric Schmidt Warns China Could Win AI Race Despite 6-Month Lag Through Cheaper, Free Models

Eric Schmidt Sounds Alarm on China's AI Strategy

Former Google CEO Eric Schmidt issued a stark warning about the AI race, stating that China trails the US by roughly 6 months in AI development but could still emerge as the winner. Speaking on the "In Good Company" podcast with Nicolai Tangen, CEO of Norway's $2 trillion sovereign wealth fund, Schmidt explained that cheaper Chinese AI models pose a significant threat to American dominance. "If they're six months behind, then we're already losing because their strategy is that the products are free," Schmidt said2

2

. China's open-weight models are priced below closed US systems, putting American products at risk of being undercut globally1

1

.

Shrinking Technology Gap in Frontier AI Model Development

The technology gap between US and Chinese AI capabilities has narrowed dramatically. Schmidt said in May that the best Chinese models sit less than 6 months behind the strongest American systems, a significant reduction from the 1-2 year gap estimated in 20251

1

. Schmidt characterized what separates the two sides as trivial, a matter of "nanoseconds"1

1

. This assessment aligns with other expert evaluations. A report from the Institute for Progress estimated that the US is roughly 6 months ahead on frontier AI models, while Demis Hassabis, co-founder of Google DeepMind, has said Chinese AI firms trail Western capabilities by around 6 months1

1

.

US Chip Export Controls Failing to Slow Chinese Progress

Schmidt warned that US chip export controls are starting to fall short of their goals1

1

. He credited China's continued strength on less powerful chips to Chinese software innovation and workarounds, citing DeepSeek as a prime example1

1

. Schmidt also noted that China gained access to advanced chips through methods that circumvented restrictions, stating they were "pilfered or smuggled" into the country2

2

. Chinese developers are leveraging distillation techniques, a method that trains one model on answers produced by a stronger one, to offset hardware shortfalls1

1

. In distillation, developers query Western AI systems and train their own models using the responses2

2

.

Low-Cost AI Models Threaten US Business Models

Chinese AI models are often free or less expensive, potentially attracting consumers in developing economies that cannot afford costly, closed Western systems2

2

. Open-source models from companies like Alibaba Qwen and DeepSeek could accelerate global adoption. IMD Business School professor Howard Yu told Benzinga that OpenAI and Anthropic's 4-6 month lead over Chinese AI rivals could prove "very perishable," as many users do not need the most advanced models2

2

. Schmidt illustrated this with an example of Tencent Holdings' WeChat agent, which quickly generated personalized guidance by scanning medicine packaging, demonstrating how less sophisticated AI can deliver practical value2

2

. For businesses weighing AI spending against hiring more employees, cost matters significantly2

2

.

Schmidt Dismisses AI Bubble Concerns Despite Market Speculation

Schmidt, who personally owns a stake in Anthropic, rejected comparisons between the AI industry and a speculative bubble2

2

. He pointed to Anthropic's revenue growth and paying customers as evidence of commercial demand, stating that Anthropic surpassed Google's revenue-growth record during his tenure by growing at more than twice the pace2

2

. "That's real growth. It's not fake. It's real money, with real customers who are paying for something useful," Schmidt said2

2

. Schmidt led Google from 2001 to 2011 and currently serves as chair and CEO of aerospace manufacturer Relativity Space1

1

. He chairs the Special Competitive Studies Project and owns about 1% of Anthropic, positioning him for a significant payout if the company goes public1

1

. Schmidt holds nearly 16% of Alphabet according to FactSet data1

1

. Companies like Nvidia-backed Reflection AI have already begun targeting low-cost Chinese models like DeepSeek with their own offerings, signaling intensifying competition in the AI race2

2

.