Text AI

China Forces Top AI Researchers to Surrender Passports

Beijing expands passport seizures to AI staff at Alibaba, DeepSeek, and other private firms, treating frontier AI talent as a strategic national security asset.

China Forces Top AI Researchers to Surrender Passports
May 26, 2026
2 min read
By Sarah Chen

Key Takeaways

  • China now requires AI researchers at private firms like Alibaba and DeepSeek to surrender their passports
  • The gap between Chinese and American AI models has narrowed to just 2.7 percent
  • Chinese AI talent migration to the United States has dropped 89 percent since 2017
  • Beijing is also blocking US investment in Chinese AI firms without government approval

China is ordering top artificial intelligence researchers at private companies to hand over their passports, Bloomberg reported today. The travel restrictions now extend well beyond state-linked firms to include senior staff at Alibaba, DeepSeek, and other leading companies building frontier AI models. The move signals that Beijing views its AI workforce as a strategic national asset too valuable to risk losing abroad.

Passports Seized Across Major AI Labs

Senior engineers and researchers at several prominent Chinese AI companies must now surrender their travel documents to employers. The official justification is that their work could expose information classified as state or commercial secrets. Companies affected include DeepSeek, whose breakthrough R1 model shook the global AI industry in early 2025, as well as Moonshot AI, StepFun, and ByteDance.

The passport seizures are not entirely new. DeepSeek staff reportedly began handing over documents in March 2025 after the company's model gained worldwide attention. But the expansion to a broader set of private-sector firms marks a significant escalation. Late last month, Beijing's National Development and Reform Commission also directed leading AI firms to reject American investment capital without prior government clearance. Together, these measures tighten control over both the people and the money behind China's AI ambitions.

A Talent War With Global Consequences

The restrictions reflect a dramatic shift in how China manages its technology sector. The capability gap between Chinese and American AI models, sometimes called large language models or LLMs, has narrowed from roughly 17 to 31 percent in mid-2023 to just 2.7 percent today. China now files nearly 70 percent of all global AI patents and produces roughly 23 percent of worldwide AI research papers.

Meanwhile, the flow of Chinese AI talent to the United States has dropped 89 percent since 2017. Beijing appears intent on accelerating that trend by making it physically harder for top researchers to leave. The approach blurs the boundary between corporate employment policy and government-directed national security controls.

For the broader AI ecosystem, the consequences could be far-reaching. International research collaboration between Chinese and Western scientists faces new barriers. Companies worldwide that rely on recruiting from China's deep talent pool may struggle to attract candidates. And Chinese AI professionals increasingly find their careers shaped by geopolitical forces well beyond their control.

Stay Informed

Weekly AI marketing insights

Join 5,000+ marketers. Unsubscribe anytime.