AI salaries soar, but China can’t find enough experts

Artificial intelligence (AI) employment and new economy jobs were a hot topic at China’s “Two Sessions” (the annual meetings of the National People’s Congress and the Chinese People’s Political Consultative Conference) this year. Outside the meetings, several leading tech companies have also launched an intense competition for talent, collectively offering nearly 30,000 positions recently, with AI-related roles taking a record share, and the average monthly salary for AI positions reportedly exceeding 60,000 RMB (US$8,625).

On 10 March, Baidu launched its largest-ever summer internship recruitment, offering more than 5,000 internship positions, over 90% directly related to AI. On 6 March, ByteDance began recruiting 7,000 interns worldwide in what it described as its largest internship-to-full-time hiring programme, with strong demand for roles in algorithm, AI engineering and AI product development.

Tencent, currently recruiting on the largest scale, plans to offer 10,000 internship positions, with technical and product roles expanding by 36% and 39% respectively, mostly focused on AI large language model development and AI product roles.

Ant Group has not disclosed its hiring scale for this year, but at least 70% of its technical positions are directly related to AI.

This latest wave of competition for AI talent among China’s major tech firms comes as the Two Sessions focus on AI and the new economy. For the first time, this year’s government work report calls for creating “new forms of smart economy” and for the third consecutive year sets out plans for “AI Plus” initiatives.

Driven by policy support and growing industry momentum, competition among companies for AI talent has intensified, with salaries rising accordingly.

A report released by China’s workplace social platform Maimai (脉脉) shows that, in the first two months of the year, the share of AI roles among all new-economy jobs has surged roughly twelvefold year-on-year. The average monthly salary for AI positions reached 60,738 RMB, about 26% higher than in other new-economy sectors.

According to the four hottest sectors in the Spring 2026 recruitment season released by Maimai on 25 February, AI is currently one of the most sought-after booming industries. Several humanoid robotics companies that featured in this year’s CCTV Spring Festival Gala are investing heavily in recruitment, with annual salaries for multiple roles exceeding 1 million RMB. Meanwhile, ByteDance’s Doubao is offering an annual salary of 1.28 million RMB for a “large language model application architecture expert”.

Yuan Yijia, co-founder of Dada Consultants, a headhunting firm specialising in top AI talent recruitment for Chinese and Singaporean tech companies, said this did not surprise her. She told Lianhe Zaobao that AI firms in cities such as Beijing and Shanghai are offering fresh PhD graduates annual salaries of around 600,000 to 1 million RMB, with top-tier talent commanding up to 2 million. “This year’s AI talent market is quite staggering — it was nothing like this a year ago.”

She also revealed that, since December last year, more than half of the Chinese clients her company has dealt with are AI-related. Unlike before, when recruitment was concentrated in major tech giants, an increasing number of startups are now competing for AI talent.

However, Yuan thinks that this is not a “talent war” in the traditional sense, but rather a result of the extreme scarcity of top-tier AI talent.

The report above showed that the supply-demand ratio for AI positions is 0.97 — that is, 0.97 people competing for each job, meaning jobs are waiting for talent. Talent is clearly more scarce here than in other new-economy sectors, where the ratio is at 1.79.

Chinese economist Ye Xiaoyang also said when interviewed that the marked increase in demand for AI talent does not mean the market has entered a booming phase across the board. It would be more precise to say that this indicated a structural shortage.

Ye added: “The most sought-after roles are a small number of high-end technical posts. These jobs command a high salary because the quality of talent supply has not yet caught up.”

Economist Pan Helin commented in an interview that the shortage of AI talent reflected that tertiary institutions, among other organisations, need to train more people specifically catered to the AI era.

Over the past two years, Chinese higher education institutions have sped up the restructuring of their degree programmes. In 2025, the Ministry of Education approved 29 new undergraduate majors, including AI education and smart audiovisual engineering, while also creating new interdisciplinary fields such as “embodied intelligence”, thereby expanding AI enrolment.

Ye said it is reasonable for universities to adjust in line with technological changes, but warned that if this is done in a simplistic fashion by “chasing hot trends”, it could bring about new structural mismatches.

He said that if institutions merely add “AI-labelled majors” without simultaneously restructuring curricula, strengthening foundational training and deepening industry placements, graduates could end up neither technically competent enough nor truly interdisciplinary. In the end, firms still would be unable to find suitable hires, and some graduates would still face employment pressure.

Employment pressure stems not only from how people are trained professionally, but also from the risk of traditional jobs being replaced by AI. As the number of university graduates hits new records, pressure on the job market would deepen.

During the Two Sessions, China’s Ministry of Human Resources and Social Security stressed that AI will be used to create new jobs and to empower traditional roles.

Ye commented that this shows there has been a major shift at the policy level vis-a-vis how the relationship between AI and employment is understood. “The government has realised that AI’s real impact on employment is not simply cutting jobs, but in reshaping job structure”.

The draft outline of China’s 15th Five-Year Plan calls for mechanisms to monitor AI’s impact on employment, strengthen job stability safeguards, expand retraining programmes, and support workers in transition.

Pan noted that while AI may reduce labour demand in some areas, it could also create jobs, and the government is considering this potential. He emphasised that more people must become AI practitioners, and human capital investments in AI-related skills must be enhanced so that the average person can adapt to the AI-era labour market. Otherwise, “widespread AI substitution could widen the wealth gap, reduce low-income purchasing power, and trigger economic contraction”.

This article was first published in Lianhe Zaobao as “中国大厂高薪抢AI人才 就业结构或面临重塑”.

Related: China’s confidence campaign: Can slogans solve economic woes? | When tech writes the story, do we still need the press?