Despite the Chinese government’s major policy push to become the world’s primary artificial intelligence (AI) innovation center by 2030, the nation is losing much of its machine learning talent to US employers, according to a Tuesday report from Diffbot.

The report identified more than 720,000 people skilled in machine learning across the globe.

While five of the top 10 global machine learning talent-producing universities are in China, their graduates are not staying, the report found. Four of these schools–Tsinghua University, Peking University, Shanghai Jiao Tong University, and the University of Science and Technology of China–have produced a total of 12,521 graduates. However, only 31% of these graduates stayed in China, while 62% left for the US, the report found.

SEE: IT leader’s guide to the future of artificial intelligence (Tech Pro Research)

Global distribution of machine learning talent is heavily centered in the US, according to the report. Nearly 221,600 workers skilled in machine learning–representing about 31% of the global talent pool–reside in the US. That means America is home to more top AI talent than the rest of the top 10 nations combined.

Following the US in terms of amount of skilled machine learning workers are India (59,980), the UK (27,425), Canada (19,259), and China (18,446).

Within the US, California dominates the machine learning workforce, with 74,791 skilled workers living there–not a surprise, given the number of tech companies located in Silicon Valley hiring this talent. California is home to more machine learning workers than the next three states on the list–New York, Texas, and Massachusetts–combined, the report found.

In terms of which employers have hired the most machine learning employees, topping the list are Microsoft (14,965), IBM (11,087), and Google (10,247). Carnegie Mellon University is the topacademic machine learning employer (3,854), with Stanford close behind (3,526).

Men and people under age 40 make up the majority of machine learning workers, the report found. About 75% of this population is male, and about 25% is female. The number of women with machine learning skills in the US is greater than the entire machine learning workforce in the other top five countries combined, according to the report.

These numbers suggest that the US is still the world leader when it comes to AI talent. However, other reports indicate that China is quickly catching up, now producing more research papers on deep learning per year than any other nation, according to the MIT Technology Review, and overtaking US companies in terms of startup funding as well. US companies will need to continue to attract top talent in this field to remain competitive.

The big takeaways for tech leaders:

  • Of the top four Chinese universities producing machine learning graduates, 62% of those graduates take jobs in the US. — Diffbot, 2018
  • Nearly 221,600 workers skilled in machine learning–representing about 31% of the global talent pool–reside in the US. — Diffbot, 2018