The ruling
Two Chinese courts have ruled that companies cannot fire employees simply to replace them with artificial intelligence. The decisions, both upheld on appeal in cases decided this spring, treat AI-driven layoffs the same way as any other termination under China’s Labor Contract Law: employers still have to prove a legally valid reason, and “we automated the role” is not one.
The Hangzhou case
The lead case involves an employee surnamed Zhou, who worked in quality assurance verifying the accuracy of outputs from large language models. Zhou’s employer tried to move her into a lower-paid role, cutting her salary from 25,000 yuan (around $3,655) to 15,000 yuan, before dismissing her. She sued, first at the Yuhang District Court and then on appeal at the Hangzhou Intermediate People’s Court, which lost the company’s case in April 2026 and ordered additional compensation.
A separate case at the Guangzhou Intermediate People’s Court reached the same result for a graphic designer whose employer had reassigned her to other duties at reduced pay before dismissing her. Courts in both cities pointed to the same detail: if a company can still find lower-paid work for an employee to do, it has effectively admitted the job still exists.
What the law requires
Under China’s Labor Contract Law, an employer can only end a contract without the employee’s consent on specific grounds: misconduct, incompetence, or a “major change in the objective circumstance” that makes the contract impossible to fulfil. Chinese courts have now found that adopting AI tools is a business decision, not an external shock, and does not meet that bar on its own.
“Employers cannot shift the risks associated with normal technological updates onto employees,” said Judge Chen Shiyuan.
The Caixin Global reporting on the rulings frames the standard the same way: AI adoption is “a controllable business strategy rather than an unavoidable disruption.” Employers are expected to retrain, reassign, or offer compensation before they can point to automation as grounds for termination.
A precedent from Beijing
The pattern isn’t new. A case decided by the Beijing Human Resources and Social Security Bureau, published in December 2025, went the same way. An employee surnamed Liu had done manual map data entry since 2009. The company switched to AI-based data collection in early 2024, eliminated her position, and dismissed her later that year. She won compensation through arbitration on similar grounds: the switch to AI didn’t excuse the employer from its normal termination obligations.
Enforcement is a separate question
Whether the rulings change behavior at scale is less clear. Eli Friedman, a labor scholar at Cornell University, was blunt about the gap between the law on paper and how it gets applied.
“The state does not have the capacity or, frankly, the interest to rigorously enforce its own laws,” Friedman said.
That leaves individual workers to bring their own cases, case by case, against employers with far more resources. The rulings set a legal standard. They don’t, on their own, create a mechanism to police it across China’s labor market.
No equivalent protection elsewhere
The contrast with the US is direct. There is no federal law requiring an employer to justify a layoff caused by automation, AI or otherwise. A proposed bill, the AI Tax and Work Protection Act, would create a Work Protection Administration funded by a tax on AI companies, but it hasn’t passed. For now, Chinese courts are setting a legal test for AI-driven layoffs that has no counterpart in the world’s other major AI market.
Operators and suppliers running AI into customer service, content generation, or back-office roles are watching the same tension play out in miniature: courts and regulators are starting to ask what an employer owes a worker whose job an algorithm can now do. China has answered that question with a specific legal standard. Other jurisdictions, including in Europe, are still working theirs out.
Source: Caixin Global
