Ángel Olmedo Jiménez
Garrigues, Madrid
angel.olmedo.jimenez@garrigues.com

Artificial intelligence has become an inescapable field having a direct impact on managing labour relations in various areas.

A few facts about the Spanish market give a clear picture of the importance of AI:

1. Spanish businesses have increased their use of AI and Big Data tools, according to a study on the use of AI and Big Data at Spanish companies.^[1]
2. In micro-companies (those with less than ten workers), AI is present in 4.6 per cent of cases, marking a one per cent increase from the 2021 figure. In large companies (above 250 employees), there has been a sharper increase, of up to eight per cent. The leading sector in the use of AI is information and communications (41.9 per cent), followed by information and communication technologies (41.3 per cent).
3. The general Spanish legislation, namely article 64.4.d of the Workers’ Statute, recognises legal representatives’ right to be informed by the employer of the parameters, rules and instructions forming the basis of algorithms or AI systems having an effect on decision-making which can have an impact on working conditions and on access to and the keeping of jobs, including profiling.

As early as in 2022, the Ministry of Labor and Social Economy assessed the possibility of making it mandatory for companies to bargain with workers’ legal representatives, ‘the variables, parameters or other characteristics of the algorithm’, especially where it is used to adopt decisions in relation to staff recruitment, shift management or adjustments to staff numbers.

In this context, EU lawmakers’ concerns on this subject are reflected in the EU Artificial Intelligence (AI) Act (approved by the legislative resolution of 13 March 2024).

Two broad focus areas are defined in that Act which can be summarised as follows: (1) the determining role of collective bargaining, especially in the rules on measures ensuring the proper use of AI; and (2) the inclusion of information obligations for workers’ legal representatives in matters related AI.

Proof that these concerns are spreading among employer and employee representatives can be seen in the fact that several Spanish collective bargaining agreements include specific rules on AI.

The countrywide collective bargaining agreement for the banking industry recognises that the use of AI can bring more efficient management of companies, although it shows a certain amount of caution over its rollout at companies.

It sets out a protection in this respect, whereby:

‘workers have a right not to be the subject of decisions based solely and exclusively on automated variables, except in the scenarios stipulated by the Law, as well as a right to non-discrimination in relation to decisions and processes, where both are based solely on algorithms, being able to request, in these cases, the assistance and participation of the individuals appointed for this purpose by the Employer, in the event of a discrepancy.’

The collective bargaining agreement further provides that the information that needs to be provided to the legal representatives must at least relate to ‘the data feeding the algorithms, the logic of their functioning and an assessment of the results’.

The countrywide collective bargaining agreement for financial credit institutions contains a similar protection to the one in the banking collective bargaining agreement, together with stipulating that companies must inform legal representatives ‘on the use of data analytics or artificial intelligence systems where decision-making processes relating to human resources and labor relations are based exclusively on digital models with no human intervention’.

Another agreed document also containing AI clauses is the countrywide agreement for the insurance industry. In addition to describing the advantages of responsible use of AI, it stipulates that it must be oriented towards ‘strengthening the capabilities and skills of workers and its use in the workplace’, although based at all times on the principle of ‘human oversight’.

The same collective bargaining agreement further underlines that the use of AI must be guided by ethical and transparent principles preventing discrimination and ensuring that information is provided to the workers’ legal representatives. Identical or similar provisions are found in companywide collective bargaining agreements.

It is clear from the above that there are currently very few rules, but that these are likely to evolve and adapt in response to the specific challenges posed by the growing use of AI.

In any event, there are certain elements, in addition to those described above, that will have to be taken into account when it comes to collective bargaining, such as (1) the potential impact on jobs after the rollout of AI and the measures available to soften its effects; (2) the cross-functional nature of AI, in that it will not simply be a labour related topic, but will also relate to data protection or occupational risk prevention, among other matters; and (3) the importance of putting training policies in place for workers that will enable them to adjust to the major changes that a field of the size we are dealing with can cause in their professional environment.