The Future of Employment in an AI Economy

Artificial intelligence (AI) is actively restructuring labour markets across sectors, from logistics and translation to legal and administrative work. As this transformation accelerates, courts are increasingly confronted with a critical legal question: can an employer lawfully dismiss an employee simply because artificial intelligence can perform their role?

Recent jurisprudence suggests a converging answer across jurisdictions: AI adoption, standing alone, does not constitute sufficient legal justification for dismissal.

China’s Hangzhou decision: A doctrinal anchor

The most direct judicial engagement with this issue has emerged from China.

In a landmark 2026 ruling by the Hangzhou Intermediate People’s Court, the dismissal of Zhou, an employee whose role had been replaced by Large Language Models (LLMs), was held unlawful. The Court, in its ruling on April 30, rejected the employer’s argument that technological advancement constituted a “major change in objective circumstances,” instead characterising AI adoption as a voluntary business decision.

Crucially, the Court focused on the employer’s failure to meet established labour law obligations:

• No attempt at retraining 

• No reassignment or redeployment efforts 

• No exploration of less drastic alternatives 

The judgment affirms that efficiency-driven technological substitution does not displace statutory protections governing termination.

This reasoning is reinforced by a 2025 determination of the Beijing Municipal Labour Bureau, which similarly held that dismissal following AI replacement was unlawful where the employer failed to act in good faith or consider redeployment.

Taken together, these decisions establish a clear principle: AI-driven substitution of labour does not, without more, satisfy the legal threshold for dismissal.

Comparative jurisprudence

Italy and Spain: Conditional legitimacy of AI-driven restructuring

European courts have adopted a more conditional approach, but one that ultimately aligns with the Chinese position.

In Italy, the Genoa Court ordered reinstatement of an employee dismissed following AI replacement, holding that the employer breached its obligation of repêchage — the duty to identify suitable alternative positions before termination. The failure to preserve employment rendered the dismissal unlawful.

However, in Court of Rome (Labour Section), Judgment No. 9135 (19 November 2025), dismissal was upheld where the employer demonstrated that:

• A genuine economic crisis drove the restructuring; 

• AI adoption formed part of a broader organisational transformation; 

• Termination was not merely a cost-reduction strategy. 

Spain reflects a similar doctrinal distinction.

The High Court of Justice of Castilla y León upheld dismissals in the translation sector where employers demonstrated a sustained decline in demand attributable to machine translation, constituting a sector-wide economic shift. By contrast, the High Court of Justice of Catalonia (2026) has invalidated dismissals in which AI-supported reasoning failed to provide an individualised and adequate justification.

European position: AI adoption may justify dismissal only where embedded within demonstrably necessary restructuring, and where obligations to preserve employment have been exhausted.

India and the UK: Procedural integrity as a safeguard

In India and the United Kingdom, courts have focused less on AI itself and more on decision-making processes.

Indian labour jurisprudence continues to emphasise human oversight and natural justice, rendering fully automated dismissals vulnerable to legal challenge. Observations from the Delhi High Court caution against unregulated reliance on algorithmic decision-making, particularly where fundamental employment rights are implicated.

Similarly, UK tribunals applying the Employment Rights Act framework have underscored the requirement of meaningful consultation. Dismissals have been deemed unfair where employees are effectively pre-selected for redundancy on the basis that their roles will be automated, thereby undermining procedural fairness.

Common thread: technological adoption cannot dilute procedural safeguards.

Africa: Doctrinal readiness without direct precedent

Although African jurisdictions have yet to produce reported decisions directly addressing AI-driven dismissal, existing labour law frameworks provide strong indicators of judicial direction.

In South Africa, such dismissals are treated as retrenchments based on operational requirements, triggering obligations of consultation, fairness, and objective selection criteria.

In Kenya, the High Court has demonstrated a broader insistence on human accountability in the use of AI, including the striking out of AI-generated pleadings, signalling judicial resistance to automation in legally significant processes.

Nigeria similarly subsumes technological workforce reductions within established redundancy doctrine, including negotiated severance and fair selection principles.

Conclusion: An emerging transnational principle

Across jurisdictions, a coherent principle is emerging:

Artificial intelligence may transform work, but it does not transform the legal standards governing dismissal.

Courts are increasingly aligned around three core requirements:

1. Substantive justification grounded in genuine operational necessity; 

2. Procedural fairness, including consultation and transparency; 

3. Last-resort termination, following reasonable efforts at retraining or redeployment. 

The Hangzhou decision is thus not an isolated development, but the leading edge of a broader doctrinal evolution, one in which technological progress is subordinated to enduring principles of fairness, accountability, and human dignity in employment law.