AI & Restructuring

Artificial intelligence is rapidly changing the way employers make workforce decisions. From recruitment and workforce planning to performance management and restructuring, AI-enabled tools are increasingly being used to analyse data, identify trends and support decision-making.

While these technologies offer significant opportunities to improve efficiency, consistency and workforce planning, they also introduce legal, regulatory and employee relations risks. Across many jurisdictions, regulators are increasingly focused on ensuring that employers remain accountable for employment decisions made with the assistance of AI, particularly where those decisions affect recruitment, promotion, performance or dismissal.

Restructuring presents particular challenges. AI may assist employers in identifying roles for redundancy, modelling organisational structures or assessing business needs, but it cannot replace the legal and procedural obligations that accompany workforce reductions. Employers must ensure that restructuring decisions remain transparent, objective and capable of explanation, with appropriate human oversight throughout the process.

This section explores how different jurisdictions are responding to the growing use of AI in employment decision-making and the practical steps employers can take to balance innovation with legal compliance.

UK

The increasing adoption of AI is prompting many UK employers to reassess workforce structures, redesign roles and identify new skills requirements. While AI may automate tasks and create opportunities for greater efficiency, it does not alter the legal framework governing restructurings and redundancies.

Employers remain responsible for establishing a genuine business rationale for organisational change, consulting where required, considering redeployment and reskilling opportunities, and ensuring that decisions are fair and free from unlawful discrimination. AI may support workforce planning, but accountability for employment decisions remains with employers.

The organisations most likely to realise the benefits of AI will be those that align technology implementation with effective workforce planning, reskilling, appropriate governance and legally compliant change management.

Read our full article here

Middle East 

Across the Middle East, governments continue to promote the adoption of artificial intelligence as part of wider economic transformation strategies, and employers are increasingly exploring AI-enabled tools across the employment lifecycle. Although dedicated regulation of AI in employment remains relatively limited, organisations remain responsible for ensuring that employment decisions comply with existing labour laws and general legal principles.

This is particularly relevant in the context of restructuring, workforce planning and performance management. Even where AI is used to support decision-making, employers should ensure that decisions remain subject to appropriate human oversight, are based on objective and verifiable criteria, and can be explained if challenged by employees or regulators.

For employers operating across the region, AI should be viewed as a decision-support tool rather than a decision-maker. Robust governance, regular review of AI outputs and clear accountability for employment decisions will be essential as organisations increasingly integrate AI into HR processes.

Read our full article here

Singapore

Singapore does not currently have employment legislation specifically regulating the use of AI in restructuring. Nevertheless, employers remain responsible for decisions supported by AI and must ensure that selection criteria for redundancy are objective, consistently applied and unrelated to discriminatory considerations. This will become increasingly important when the Workplace Fairness Act comes into force, as its protections will apply to adverse employment decisions, including dismissal.

AI may assist with workforce modelling, role comparisons and the identification of potential efficiencies, but employers should avoid relying on its outputs without meaningful human review. Any employee data used should also be handled in accordance with the Personal Data Protection Act. 

Employers should therefore understand the basis of AI-generated recommendations, test for potentially biased outcomes and retain a clear, independently supportable rationale for the final restructuring decisions.

The use of AI in Hong Kong is increasing across industries. Surveys indicate that over 60% of organizations are already using AI, representing a significant increase from the previous year. While AI can support restructuring, employers remain legally accountable for employment decisions. Hong Kong’s Employment Ordinance (Cap. 57) continues to govern redundancy, severance payments, notice requirements, and termination procedures, meaning employers must ensure that restructuring decisions are transparent, fair, and compliant even when assisted by AI.

Under the Personal Data (Privacy) Ordinance (Cap. 486), organizations must comply with the Data Protection Principles when collecting and processing employee data. The PCPD’s Artificial Intelligence: Model Personal Data Protection Framework, issued in 2024, emphasizes data minimization, purpose limitation, and data accuracy. Employers using AI in restructuring should conduct risk assessments and ensure that decisions are supported by accurate and lawfully collected data.

In addition, Hong Kong’s anti-discrimination laws covering sex, disability, family status, and race apply equally to AI-assisted decision-making. AI systems may unintentionally replicate historical biases. Therefore, transparency around AI usage and senior-level review of key decisions is essential to help organizations with legal compliance.

Hong Kong

The use of AI in Hong Kong is increasing across industries. Surveys indicate that over 60% of organizations are already using AI, representing a significant increase from the previous year. While AI can support restructuring, employers remain legally accountable for employment decisions. Hong Kong’s Employment Ordinance (Cap. 57) continues to govern redundancy, severance payments, notice requirements, and termination procedures, meaning employers must ensure that restructuring decisions are transparent, fair, and compliant even when assisted by AI.

Under the Personal Data (Privacy) Ordinance (Cap. 486), organizations must comply with the Data Protection Principles when collecting and processing employee data. The PCPD’s Artificial Intelligence: Model Personal Data Protection Framework, issued in 2024, emphasizes data minimization, purpose limitation, and data accuracy. Employers using AI in restructuring should conduct risk assessments and ensure that decisions are supported by accurate and lawfully collected data.

In addition, Hong Kong’s anti-discrimination laws covering sex, disability, family status, and race apply equally to AI-assisted decision-making. AI systems may unintentionally replicate historical biases. Therefore, transparency around AI usage and senior-level review of key decisions is essential to help organizations with legal compliance.

Read our full article here

Australia

In the Australian employment landscape, artificial intelligence is increasingly being used by applicants to prepare unfair dismissal and general protections claims. The Fair Work Commission (FWC) has observed a significant increase in lodgements in recent years, with the growing accessibility and speed of AI tools enabling individuals to generate claim materials with minimal time and cost.

A key concern identified by the FWC is that AI-generated content is often being submitted without adequate review. In some cases, claims contain factual inaccuracies, irrelevant legal arguments, or hallucinated references to legislation and case law. This can create inefficiencies for all parties involved as additional time and resources are being spent responding to claims that are poorly founded or based on incorrect legal information.

In response, the FWC has released a draft Guidance Note outlining expectations for the use of AI in proceedings before the Commission. The proposed requirements include disclosing the use of AI in the preparation of documents, verifying that all factual and legal content is accurate and relevant, and confirming that the material reflects the party’s own knowledge.

The proposed guidance reflects a broader regulatory focus on the responsible use of AI in legal and workplace processes. It serves as a reminder that while AI can improve efficiency, responsible use is critical to ensuring accuracy.

Chile

Chile has no legislation in force regulating the use of artificial intelligence in employment decisions. The bill on AI systems (Bulletin No. 16,821-19) is currently under parliamentary discussion and adopts a risk-based approach inspired by the European model, expressly identifying uses of AI in the workplace among those classified as high-risk; it is not yet law. Law No. 21,719 on personal data protection will be, however, entering into force on 1 December 2026 and imposing a duty to disclose the existence of automated decision-making and requiring impact assessments for processing based on automated decisions or the systematic and comprehensive evaluation of personal aspects.

In the meantime, the general employment law framework applies across the full employment lifecycle. AI-assisted recruitment decisions are subject to non-discrimination requirements, whose list of suspect grounds was extended by the Karin Act to any reason whatsoever. Performance assessments underpinning disciplinary measures must be objective and verifiable. Restructuring presents particular demands: termination on grounds of business necessity requires a cause that is objective, documented and capable of explanation before a court, regardless of the analytical tools used to identify it. Employers cannot substitute that obligation of justification with the internal logic of an algorithmic model.

This is an emerging area. For the multinational employer, the concrete risk is that any selection, performance or dismissal decision whose rationale cannot be explained in terms intelligible to a Chilean court will be indefensible, however sophisticated the underlying model. It is advisable to document the human reasoning behind each decision, and not merely the output of the system. A further consideration is that where an AI system operates outside Chile or employee data is processed by a parent company or external provider abroad, the operation will constitute an international data transfer subject to Law No. 21,719, which requires either an adequacy determination by the Personal Data Protection Agency, contractual clauses, or binding corporate rules providing adequate safeguards. 

Egypt

The Egyptian government has made AI adoption a pillar of its economic transformation agenda, most notably through the National Council for Artificial Intelligence (NCAI), which oversees the second edition of Egypt's National AI Strategy (2025–2030) and is currently steering the early stages of a dedicated AI Law. 

Egyptian employers, particularly in sectors with a strong multinational or outsourcing presence, are increasingly piloting AI-enabled tools across recruitment, performance management and workforce planning. However, dedicated regulation of AI specifically in the employment context remains at an early stage, meaning organizations must continue to test AI-assisted employment decisions against the existing framework under Labour Law No. 14 of 2025 and general principles of Egyptian civil law.

This is particularly relevant where AI tools inform restructuring, workforce planning or performance-related decisions. Even where such tools support the process, employers should ensure that final decisions remain subject to meaningful human review, rest on objective and verifiable criteria, and can be justified if challenged before the relevant authorities especially given the new Law’s heightened documentation expectations and its protections against arbitrary dismissal.

For employers operating in Egypt, AI should be treated as a support tool rather than a decision-making tool. 
As the NCAI's regulatory work progresses and enforcement activity under the new labour law ramps up, robust governance, regular review of AI outputs and clear accountability for employment decisions will be essential to managing legal risk.

Kenya

Kenya does not have legislation regulating artificial intelligence or specific applications of it, but existing laws guide adoption and use. Kenya’s primary artificial intelligence framework is guided by the Kenya AI and other Emerging Technologies Policy, 2026 and the broader Kenya National Artificial Intelligence Strategy 2025–2030.

The 2007 Employment Act adopts a broad definition of redundancy hence the framework allows for the replacement of employees if there are genuine reasons and the proper procedure is followed. Where AI is used for role restructuring, employers must obtain express employee agreement to the changes in terms. Additionally, meaningful human review remains a requirement under the Data Protection Act. As such, while AI can be used to guide decision making, it should not be the final decision maker, especially while it is well known that current systems retain biases that can be deemed discriminatory, contrary to the Employment Act.

Kenya is moving towards formal AI regulation, as shown by the 2026 Artificial Intelligence Bill. The latest draft sets up the administrative structures to regulate AI systems and classifies AI systems used in employment as high-risk and subject to several compliance obligations. Notably, the Bill calls for human-centric AI and requires systems that are likely to affect employment to be subjected to workforce impact assessments and to implement mitigation procedures to minimise harm. This will be an important consideration when restructuring in the future, as the Bill could make it procedurally or substantively difficult to replace human workers with AI systems.

Tanzania

Tanzania does not yet have dedicated legislation governing the use of artificial intelligence in employment decision-making, and restructuring remains firmly anchored in the established statutory framework under the Employment and Labour Relations Act. Where AI tools support workforce planning or help identify roles for redundancy, employers remain fully responsible for ensuring the underlying process satisfies the Act's requirements on fair reason, consultation and selection criteria.

This is a live issue for employers with regional or global HR analytics feeding into Tanzania operations. A redundancy selection process that cannot be explained in plain terms to an employee, a trade union or the CMA carries real risk, regardless of how sophisticated the underlying model was.

Employers should treat AI as an input to, not a substitute for, the judgement of the person accountable for the decision, and should keep clear records showing how AI-generated recommendations were reviewed and tested before being acted upon.

Read our full article on managing AI-assisted restructuring decisions within Tanzania's employment law framework.

Takeaways 

• Keep humans accountable: AI should inform employment decisions, not replace human judgement.
• Establish AI governance: Put in place clear governance, oversight and accountability for the use of AI in HR processes.
• Test for bias and accuracy: Regularly review AI tools and outputs to identify unintended bias or inconsistent outcomes.
• Ensure transparency: Be able to explain how AI-supported decisions were reached and the factors taken into account.
• Comply with local employment laws: AI does not remove obligations relating to consultation, procedural fairness, discrimination or dismissal.
• Maintain clear records: Document the role AI played in decision-making, together with the human review and rationale supporting the final outcome.