Chinese semiconductor and artificial intelligence firms are deploying aggressive equity incentive plans to retain critical talent amid intense domestic poaching and the U.S.-China tech race. Companies like AMEC and Cambricon Technologies have extended stock grants to over 85% of their workforces, while ByteDance and Tencent offer salary increases of up to 150%. This talent battle occurs as Chinese AI startups, heavily rooted in Tsinghua University research, rapidly close the capability gap with U.S. leaders despite facing severe high-end chip shortages and accusations of model distillation.
Chinese AI talent retention strategies
- ▪Zhongji InnoLight allocated 2.48 million shares to 99 key personnel, yielding an average value exceeding 26 million yuan per person based on its April announcement day stock price.
- ▪ByteDance and Tencent offered pay increases of up to 150% and bonuses of around 35% to secure artificial intelligence talent.
- ▪AI chipmaker Cambricon Technologies unlocked approximately 600,000 shares for 124 core staff, yielding an average stock value of 5.57 million yuan (US$828,000) per person.
- ▪Semiconductor equipment maker Advanced Micro-Fabrication Equipment China (AMEC) unveiled a restricted stock plan in March covering more than 97% of its total workforce.
- ▪Cambricon Technologies granted 5 million shares to 944 employees, covering 85.3% of its total workforce, under an incentive plan running through 2028.
- ▪Cambricon Technologies tied its employee incentive plan to a corporate revenue target of approximately $14.8 billion.
- ▪Chinese artificial intelligence and semiconductor companies are deploying aggressive equity incentive plans, including zero-cost share awards and near-blanket workforce coverage, to retain key personnel.
Tsinghua University AI research origins
- ▪Tsinghua University alumnus Yang Zhilin founded the artificial intelligence startup Moonshot AI, also known as Dark Side of the Moon.
- ▪Andrew Yao, a former Princeton professor and Turing Award winner, established the elite 'Yao class' at Tsinghua University in 2005 to cultivate top-tier computer science talent.
- ▪In 2018, the Chinese government cleared the way for researchers at state institutes and universities to start companies and market their technology.
- ▪Tsinghua University professor Tang Jie co-founded Z.AI, an artificial intelligence startup spun out of his university lab in 2019.
China-US AI capability gap
- ▪DeepSeek founder Liang Wenfeng stated that computer chips, rather than people, represent the most critical gap between Chinese and American artificial intelligence companies.
- ▪Z.AI released its GLM-5.3 model, claiming it matched Anthropic's Mythos 5 model in cybersecurity capabilities.
- ▪Industry leaders in China and the United States estimate that the capability of China's best artificial intelligence model is just months behind the best U.S. models.
- ▪Z.AI reached an annual recurring revenue of $1 billion by July, which trails Anthropic's annual revenue of $65 billion in the same month.
- ▪Researchers at Chinese labs, including Alibaba and Z.AI, are often allotted one-fifth of the high-end computer chips available to peers at OpenAI and Google.
Model distillation controversy
- ▪Anthropic accused Chinese artificial intelligence companies, including Z.AI and Moonshot AI, of violating its policies by engaging in large-scale model distillation.
- ▪Model distillation is a process where a new artificial intelligence system learns from an existing one by asking it hundreds of thousands of questions and analyzing the answers.
Global semiconductor talent competition
- ▪The European Chips Skills Academy estimated Europe's semiconductor talent gap at 65,000 workers.
- ▪Unlike Chinese firms using equity and U.S. firms using cash, European semiconductor employers struggle to use equity schemes due to regulatory differences between member states.
Story comments
Loading comments…