AI Is Already Being Used to Justify Redundancies
Employers across UK industries are restructuring workforces in response to artificial intelligence tools that automate functions previously requiring human workers. Legal document review, financial analysis, customer service, data processing, accounting, content production, and administrative roles are all affected. Employers are increasingly citing "AI implementation," "digital transformation," or "automation" as the business reason for redundancy. In 2025 and 2026, over 77,000 technology sector roles globally were eliminated with AI cited as a contributing factor.
The critical point for employees: an AI business reason does not create an AI exemption from the legal requirements for a fair redundancy process. Every legal protection that applies to any redundancy applies equally when the stated reason is AI or automation. Some protections become more important, not less.
What Makes an AI-Driven Redundancy Genuine
A genuine redundancy exists where the requirement for employees to do work of a particular kind has ceased or diminished. If an employer has implemented an AI system that genuinely performs the functions of a role and no longer requires a human to perform them, that is capable of being a genuine redundancy situation. The employer does not need to prove the business will fail without the change -- only that the requirement for that work has genuinely reduced.
However, "AI is doing your job now" is not a self-executing justification. The employer must still be able to demonstrate that the AI system actually performs the relevant functions, that the reduction in headcount is genuinely connected to the implementation, and that the decision to eliminate roles was made for business reasons rather than to target particular individuals under cover of a technology change.
Where AI Redundancy Becomes Unfair
An AI-driven redundancy is unfair where the consultation process was inadequate. Employers using AI restructuring programmes often move quickly -- the technology is implemented, the headcount change is announced, and employees are out within weeks. Speed does not excuse the obligation to consult genuinely, to consider alternatives including retraining, and to look for suitable alternative vacancies. A fast-moving AI implementation programme that does not allow for proper consultation produces unfair dismissals regardless of the legitimacy of the underlying business reason.
It is also unfair where the selection process was flawed. If an employer uses an algorithm or an AI system to assist in identifying which roles or employees to eliminate, the selection must still be fair and objective. Ask specifically: was any automated system used in the selection process? How was it designed? What criteria did it apply? You have the right to challenge a selection that was influenced by an automated system, particularly if you cannot understand how you came to be selected.
The AI Selection Bias Risk
AI systems trained on historical data can replicate and amplify historical biases. An AI tool that uses historical performance data, promotion patterns, or role descriptions to identify redundant positions may systematically disadvantage certain groups -- older workers, women, workers from particular ethnic backgrounds -- if those patterns are embedded in the training data. This is not a hypothetical concern. It has already been identified in hiring algorithms used in the US and is equally applicable to redundancy selection tools.
If you are in a group that is disproportionately affected by an AI-assisted selection process, consider whether the outcome constitutes indirect discrimination. Indirect discrimination occurs where a provision, criterion, or practice that appears neutral in fact puts people sharing a protected characteristic at a particular disadvantage. An AI selection tool that disproportionately eliminates older workers is potentially applying a practice that amounts to indirect age discrimination -- even if age was not explicitly used as a criterion.
The Retraining Question
Before completing any redundancy, your employer must consider whether suitable alternative vacancies exist and whether you could fill them. In an AI-driven restructuring this question takes on particular importance. If the employer is implementing AI tools that require human oversight, management, or augmentation -- roles that did not previously exist -- those are potential suitable alternative vacancies. If younger or more recently hired employees are being offered retraining and transition into new AI-adjacent roles while older or longer-serving employees are selected for redundancy, that pattern may constitute age discrimination.
Ask specifically in writing: are any new roles being created in connection with the AI implementation? If so, have you been considered for them? The answer to this question may determine whether the redundancy is fair.
What to Do If You Have Been Made Redundant Because of AI
Ask in writing for the full business case for the AI implementation and how it connects to your role. Ask whether any automated or algorithmic system was used in the selection process and if so, what criteria it applied. Ask what retraining or alternative roles were considered for you specifically. Ask for your selection scores and the methodology used. Raise any concern that the selection may have disproportionately affected employees sharing a protected characteristic. Then contact ACAS at acas.org.uk within three months less one day of your dismissal date and start early conciliation. The redundancy may be unfair, the selection may be discriminatory, or both. A failure to consult properly is a separate ground. Multiple claims can be brought simultaneously.
The ERA 2025 Protective Award
If 20 or more employees have been made redundant in connection with the AI implementation without proper collective consultation, each affected employee may be entitled to a protective award of up to 180 days pay (from 6 April 2026 under ERA 2025, doubled from the previous 90-day maximum). This award is separate from unfair dismissal compensation and is available even where the underlying business reason for the redundancy was genuine. Many AI-driven restructuring programmes move too quickly to comply with collective consultation requirements. If yours did, the protective award claim is worth pursuing.