May 2026
The emergence of AI is a reality. However, the speed at which it is permeating every layer of social life is exceeding all expectations.
There are multiple dimensions in which AI is generating highly engaging debates. Some commentators even place the discussion on a level that often goes unnoticed: while society concerns itself with debating whether an algorithm should have a “bias” of 2% or 4%, the ethical limits of AI, or whether a self-driving car should run over an elderly woman or a child when forced to choose (the classic trolley dilemma), we may have relinquished our most fundamental sovereignty as human beings: deciding whether we want this new technology at all.
In the field of employment relations, transformation appears inevitable. The law, which is traditionally slow to adapt, will have to respond to a foreseeable and relatively immediate technological substitution across various sectors. The key will lie in a critical reassessment of the legal frameworks available in each jurisdiction.
In Spain, legal doctrine concerning dismissals based on automation, algorithms or Artificial Intelligence has not yet been fully consolidated as an autonomous line of case law, particularly with regard to generative AI, management algorithms or automated decision-making systems. To date, the courts have not established an independent category of “AI dismissal” or “automation-related dismissal”.
Instead, the issue has been addressed within the ordinary framework of objective dismissal under Article 52(c) of the Workers’ Statute, in conjunction with the grounds set out in Article 51.1 thereof: economic, technical, organisational or production-related reasons.
The central principles are as follows:
- The introduction of technology may be legally relevant as a technical ground, and often also as an organisational one.
- However, the employer must demonstrate a genuine cause, a functional link between the introduced technology and the redundancy of the role, and the reasonable appropriateness of the termination measure.
- Mere gains in efficiency, cost reductions, or the abstract replacement of human labour by software are not, in themselves, sufficient unless there is evidence of a real impact on labour needs and on the specific position being eliminated.
Freedom of enterprise and the pursuit of productivity improvements are legitimate considerations. Nevertheless, the courts do not accept that any technological measure that reduces costs automatically justifies an objective dismissal.
The prevailing criterion in general employment case law, equally applicable to automation, is that the measure must be reasonable from a business management perspective and not arbitrary, capricious, or disconnected from the productive organisation.
At present, there is no statutory obligation on employers to redeploy affected employees. This may give an unexpected prominence to collective bargaining, which is capable of establishing new rights and obligations, particularly in cases of large-scale impacts in sectors exposed to the effects of AI development. Measures such as training and reskilling, technological adaptation, safeguards against biased algorithmic decisions, and protection for vulnerable groups facing the digital divide could come to occupy a central place in trade union agendas.
An analysis of current case law provides insight into the factors that lead to one or another judicial outcome. For instance, a finding of unfair dismissal is often associated with a generic dismissal letter that fails to identify the technology introduced, does not explain the affected tasks, does not quantify the reduction in workload, or does not justify the selection of the employee concerned. Dismissal will likewise be deemed unfair when the tasks continue to be carried out by other employees without evidence of a real reduction, or where the alleged technical ground merely conceals a cost-cutting measure insufficiently linked to the productive organisation.
Accordingly, the proper identification and compilation of relevant data, the establishment of a functional connection with the elimination of the role, and the reasonableness of the employer’s decision (all of which could, incidentally, be facilitated by AI, albeit under the necessary guidance and supervision of legal counsel) must underpin the narrative set out in the dismissal letter if it is to be upheld as fair.
It is not unreasonable to anticipate legislative or judicial intervention in this area should the projected levels of social and business impact materialise. In the meantime, it falls to the various stakeholders and legal practitioners to anticipate solutions capable of integrating new business needs into the existing regulatory framework, in other words, to envisage new spaces for action where none previously existed, something that no AI can, for the time being, promise.
Partner
