Cross-Border AI Compliance Risks Unveiled by Meta’s Manus Deal
Meta’s $2 billion acquisition of AI startup Manus has become a cautionary tale for enterprise tech buyers. Despite Manus relocating from Beijing to Singapore in 2025, Chinese regulators are scrutinizing whether the deal violates export controls and other laws. This situation highlights an important point: where a company is registered doesn’t necessarily tell you about its regulatory exposure. For businesses investing in AI, understanding the technology’s regulatory landscape is crucial.
The Reality Behind Corporate Domicile and Regulatory Risks
Although Manus moved its 105-person team from Beijing to Singapore, Chinese authorities are still investigating the deal. The Ministry of Commerce announced on January 9 that it would assess if the acquisition breaches export controls, technology transfer rules, or investment regulations. This underscores a key truth: the jurisdiction of a company’s registration isn’t the sole factor in determining legal compliance. Instead, the focus is on the actual technology and how it’s transferred or used across borders.
Experts like Dai Menghao, a partner at King & Wood Mallesons specializing in export controls, explained that Chinese regulators could still target Manus’s AI technology, regardless of where the company is registered. The technology itself, not the corporate address, determines which laws apply. This means that relocating a team or changing corporate structures might not shield a company from legal risks related to technology exports or transfers.
Relocation Doesn’t Guarantee Regulatory Freedom
Manus appeared to take steps to distance itself from Chinese regulations. The company laid off most of its Beijing staff, moved operations to Singapore, Tokyo, and San Francisco, and secured US funding. Meta also stated that it would have no Chinese ownership interests after the deal and that Manus would cease its services in China. Despite these steps, Chinese regulators made it clear that compliance isn’t just about company registration. The legal obligations depend heavily on the actual transfer or export of technology and data.
He Yadong, a spokesperson for China’s Ministry of Commerce, emphasized that foreign investment and technology exports must follow Chinese laws, regardless of where a company is headquartered. The investigation will focus on whether Manus transferred technology or talent from its Chinese entities without proper licenses. If regulators find that licenses were needed but not obtained, the company’s founders could face criminal charges under Chinese law.
This situation reveals a broader issue: legal compliance in cross-border AI deals depends on the specifics of technology transfer, not just corporate structure. Companies must understand that relocating operations doesn’t automatically remove regulatory obligations, especially when dealing with strategic technologies.
Changing Export Controls and Increasing Government Oversight
China updated its technology export control rules in 2020, broadening the scope to include certain algorithms and AI technologies. These changes were widely seen as giving Beijing stronger legal tools to control outbound tech transfers. The move came after the US pressured ByteDance to sell TikTok’s US operations, prompting China to reinforce its authority over outbound technology deals.
For enterprise buyers, this means that the regulatory environment is becoming more complex. Companies involved in developing or transferring AI technologies need to be aware of evolving laws and ensure they have the right licenses before moving technology abroad. Failure to do so could lead to legal penalties or even criminal charges for founders.
Ultimately, this case shows that cross-border compliance isn’t just about where a company is based. It’s about understanding the actual technology involved and ensuring all legal requirements are met before transferring or exporting it. For enterprise AI buyers, staying informed and compliant is more important than ever in today’s global regulatory landscape.















What do you think?
It is nice to know your opinion. Leave a comment.