MUMBAI: If machines are taking over jobs, courts in China are now making sure they don’t take over accountability too. A series of recent rulings across China is reshaping how companies deploy artificial intelligence in the workplace, with courts making it clear that automation, by itself, is not a valid ground for terminating employees. Decisions delivered between December 2025 and April 2026 in Hangzhou and Beijing mark a notable shift in judicial thinking from viewing AI as an unavoidable disruption to treating it as a conscious business choice with legal consequences.
At the heart of the rulings is a reinterpretation of China’s Labour Contract Law. Courts have clarified that adopting AI does not qualify as a “major change in objective circumstances,” a threshold typically reserved for external shocks such as mergers or relocations. In other words, if companies choose automation, they must also own its fallout.
One case in Hangzhou brought this principle into sharp focus. An employee working at the human–AI interface reviewing AI-generated outputs saw parts of his role automated as systems improved. The company reassigned him to a lower role with a 40 per cent pay cut. When he refused, his contract was terminated with compensation of just over 311,000 yuan.
The employee challenged the move and won at every level, including before the Hangzhou Intermediate People’s Court. Judges ruled the dismissal unlawful, stating that partial automation does not make a role redundant if core responsibilities remain. The reassignment was also deemed unreasonable for failing to preserve both pay and professional standing.
A similar line was taken in Beijing, where an employee dismissed after automation of map data collection successfully challenged his termination. Authorities reiterated that companies cannot use AI adoption as a workaround to bypass labour protections.
Taken together, the rulings establish a clearer standard: employers must prove that a job has become genuinely unviable not merely more efficient to automate. They are also expected to explore redeployment, offer retraining and engage in meaningful consultation. Failure to do so could lead to reinstatement orders or higher severance payouts.
The implications stretch beyond China. While jurisdictions such as the US and Europe have largely relied on advisory frameworks for AI, Chinese courts are stepping in directly, shaping how automation interacts with employment law in real time.
The ripple effects are already visible elsewhere. In India, courts have begun pushing back on unchecked AI-led decisions. The Madras High Court in 2026 set aside an AI-driven termination, while the Delhi High Court has cautioned against excessive reliance on automated systems.
Existing Indian laws, including the Industrial Disputes Act, treat job losses due to automation as retrenchment requiring notice, compensation and due process. The Digital Personal Data Protection Act, 2023 adds another layer by regulating how employee data can be used in AI systems.
With a proposed Artificial Intelligence (Ethics and Accountability) Bill, 2025 also in the works, the direction of travel is becoming clearer. As AI reshapes workplaces, courts across jurisdictions are signalling the same message: efficiency may be automated, but fairness cannot be.
