发生了什么
Chinese authorities have rolled out a prior‑approval regime for short overseas trips by core personnel in private AI and semiconductor companies, extending the requirement to their spouses and children. The system, reported by Bloomberg and summarized by Chosunbiz, applies to executives and researchers at firms such as Alibaba and DeepSeek. The list of covered individuals is compiled by State Council ministries, notably the Ministry of Commerce, which assesses national importance. The move follows a May directive that already required AI founders and senior staff to seek approval before departure. The policy aims to prevent talent loss and potential defection amid heightened geopolitical competition.
According to Chosunbiz, which cited a Bloomberg roundup from September 28, Chinese authorities have instituted a prior‑approval system for short overseas trips targeting core personnel in private AI and semiconductor firms deemed essential to national security. The requirement now includes immediate family members—spouses and children—of these personnel.
The approval process is expected to be administered by State Council ministries such as the Ministry of Commerce, which will unilaterally assess the national importance of each individual. Bloomberg reported that several people have already been instructed to obtain prior approval for family travel, and officials plan to broaden the list over time.
The policy builds on a May directive that forced AI founders, researchers, and executives at private companies like Alibaba and DeepSeek to seek government clearance before departing China. Previously, some private AI engineers only needed to report travel plans; now they must receive explicit permission to board a plane.
Experts cited in the article link the move to concerns over talent loss after Meta’s $2 billion acquisition of the Chinese‑founded AI startup Manus. Chinese authorities reportedly pressured to nullify the deal and barred the co‑founders from leaving the country.
为什么这很重要
The policy marks a shift from typical export‑control measures that target technology, capital, and data to directly restricting the physical movement of people who possess AI expertise. By tying families to the approval process, Beijing seeks to deter brain drain and limit the risk of strategic knowledge transfer to foreign entities, especially after high‑profile acquisitions such as Meta’s purchase of the Chinese‑founded AI startup Manus. The measure also signals a broader “AI sovereignization” strategy, where talent is treated as a national security asset. While it may temporarily retain researchers within China, analysts warn it could erode trust, increase incentives for covert exits, and ultimately hamper the long‑term development of China’s AI ecosystem.
By extending travel controls to families, Beijing treats AI expertise as a strategic resource comparable to semiconductors or quantum technologies, but with a more direct human‑centric approach. This reflects a broader “AI sovereignization” effort that goes beyond traditional export controls.
The policy could temporarily curb the immediate outflow of top talent, but analysts warn it may backfire by creating a climate of distrust. Henry Gao, a law professor at Singapore Management University, told Deutsche Welle that “the more you block exits and capital flows, the stronger the incentives to find other escape routes,” suggesting a potential acceleration of talent flight over the medium to long term.
If the restrictions become pervasive, multinational AI firms with Chinese operations may need to adjust talent deployment strategies, possibly relocating research teams or negotiating exemptions, which could fragment the global AI talent pool.
互动机制：它实际上是如何运作的
以交互方式探索这一发展背后的基础技术。
Impossibility results in algorithmic fairness (e.g. Kleinberg et al., Chouldechova) show what?
接下来看什么
Future expansions of the approval list, enforcement consistency across regions, and any legal challenges or diplomatic pushback from affected companies. Monitoring how multinational AI firms with Chinese operations respond—whether they adjust talent deployment, negotiate exemptions, or relocate R&D—will indicate the policy’s practical impact. Additionally, reactions from foreign governments and the potential for reciprocal travel restrictions could shape the broader AI talent war.
Whether the approval list will be expanded to include additional companies, research institutions, or broader categories of AI work.
The consistency of enforcement across provinces and the potential for legal challenges by affected individuals or firms.
Responses from foreign governments, especially those that host Chinese AI talent, which could lead to reciprocal travel restrictions or diplomatic negotiations.
The impact on foreign investment decisions and talent recruitment strategies of multinational AI companies operating in China.