China’s DeepSeek Imposes Anti-Poaching Condition in First External Fundraise

Chinese artificial intelligence startup DeepSeek reportedly imposed an unusual condition during its first external fundraising effort: investors must agree not to recruit the company’s employees or encourage them to establish competing ventures.

According to a report published by a fundraising-focused media platform affiliated with 36Kr, founder Liang Wenfeng outlined the requirement during a four-hour virtual meeting with potential investors in May. Participants were reportedly informed that any investment in the company would be contingent upon a commitment to refrain from poaching DeepSeek personnel. The report has not been independently verified.

The condition highlights the increasingly intense competition for AI talent across China’s technology sector, where major companies are racing to develop advanced artificial intelligence systems and pursue the longer-term goal of artificial general intelligence (AGI).

DeepSeek is said to have completed its inaugural fundraising round this week, securing a valuation exceeding $50 billion and becoming the most highly valued standalone AI startup in China. Since its establishment, the company had largely avoided outside capital in order to prioritize research and development over commercial expansion. However, recent departures of several prominent researchers reportedly contributed to the decision to seek external funding.

Among those who left was Luo Fuli, a key contributor to DeepSeek’s V3 model. She joined Xiaomi late last year to lead the company’s MiMo AI initiative. Since then, Xiaomi has introduced models that have surpassed DeepSeek’s offerings on a number of performance benchmarks.

Competition for specialized AI expertise has intensified across the industry. Reports earlier this year indicated that Tencent recruited two significant AI developers from ByteDance. More recently, Tencent was also reported to have invested $20 million in a newly established AI laboratory founded by Juyang Lin, the former lead researcher behind Alibaba’s Qwen language models.

Lin announced his departure from Alibaba’s Qwen project in March. Separately, Alibaba has continued to adjust its AI strategy, including leadership changes within its enterprise-focused DingTalk division following internal discussions regarding the unit’s role in the company’s broader artificial intelligence ambitions.

Chinese technology firms are also increasingly targeting talent with experience at leading international AI laboratories. Tencent recruited Yao Shunyu from OpenAI last year to serve as its chief AI scientist. Both Yao and Liang have publicly emphasized the importance of pursuing AGI development, reflecting a growing belief among Chinese AI leaders that achieving human-level or superior artificial intelligence capabilities should remain a central industry objective.

DeepSeek, Tencent, Alibaba, and ByteDance have not publicly commented on the reported developments.