Microsoft offers relocation to hundreds of China-based AI staff amid U.S.-China tech tensions

• Sources told the WSJ that relocation has been offered to about 700 to 800 China-based employees working in machine learning and other work related to cloud computing.
• The move from Microsoft comes as Washington seeks to crack down on Beijing’s access to advanced AI and cloud computing technology.

Microsoft has reportedly asked China-based cloud computing and artificial intelligence operations employees to consider relocating out of the country, as Washington cracks down on Beijing’s access to the advanced technology. 

The Wall Street Journal broke the story on Thursday, reporting that the staff, mostly comprising Chinese engineers, had been offered the opportunity to transfer to countries including the U.S., Ireland, Australia, and New Zealand, according to unnamed sources. 

One source told WSJ that Microsoft had made the offer to about 700 to 800 people in total who were involved in machine learning and other work related to cloud computing. 

CNBC could not independently verify the report.

In a statement shared with CNBC, a Microsoft spokesperson confirmed that the company had “shared an optional internal transfer opportunity with a subset of employees” without supplying details on the number and affiliation of staff affected.

“We remain committed to the region and will continue to operate in this and other markets where we have a presence,” the spokesperson said, adding that the potential transfers would not impact operations.

Microsoft employs roughly 7,000 engineers for its Asia-Pacific research-and-development group, with most of this workforce based in China, the WSJ reports.

The move comes amid U.S. efforts to prevent China from developing cutting-edge AI technology, which could be used for military purposes. In the past two years, the U.S. has placed waves of restrictions on China limiting its ability to buy advanced chips and chip-making equipment that can be deployed to train AI models. 

Now, the Biden administration is looking to place new guardrails on the export of advanced AI models, such as the large language model that powers Microsoft-backed ChatGPT, according to recent reports. 

There is currently little government oversight stopping companies like Microsoft, one of the U.S.’s largest cloud-computing and AI players, from selling or offering AI model services to foreign entities. 

The U.S. reportedly fears that AI models, which mine vast amounts of data to generate content, could be used for cyber attacks or to create biological weapons.

Earlier this year, Microsoft released a report stating that state-backed hackers from Russia, China, and Iran had been using tools from OpenAI to hone their skills and support their hacking campaigns. 

Microsoft has been deeply ingrained in China for more than three decades, even as other Western tech companies were pushed out by strict regulation. The company says that China is home to its largest R&D center outside of the U.S.

Read the full report from Wall Street Journal.