Today, the technological revolution has led companies to integrate artificial intelligence (AI) tools in various areas, including labour relations management. However, the use of AI tools, such as ChatGPT, in processes as sensitive as dismissal, which is the most drastic decision a company can take leading to the termination of the employment relationship, poses significant legal and ethical challenges.
The recent ruling of the High Court of Justice of Catalonia of 16 October 2024 (Rec. 2759/2024), which invalidated a dismissal due to the lack of justification in a letter drafted through the use, in the words of the Court itself, of "the misnamed" Artificial Intelligence (AI), has raised several questions about the role of this technology in labour decision-making.
The article analyses the legal implications and risks of the use of artificial intelligence in labour dismissals, focusing on the Spanish and European regulatory framework, with special reference to the European Regulation on Artificial Intelligence (RIA). It highlights the need for preventive measures and good practices to ensure a responsible use of these tools in labour relations.
One of the key cases mentioned is a judgment of the High Court of Justice of Catalonia, where a dismissal was declared null and void because the letter, generated with AI, contained errors and lacked adequate grounds. It showed that the company did not comply with the legal requirements for notification of dismissal, which led to the nullity for violation of the right not to suffer discrimination, as the worker had recently been in a process of temporary disability.
It points out that the lack of human supervision in automated business decisions can lead to discrimination and unfair dismissals. Mention is made of Regulation (EU) 2024/1689, which classifies AI applications according to their level of risk and sets strict requirements for those considered high risk, such as those used in recruitment and dismissals.
Good practices to mitigate these risks are presented, including:
- Human oversight in document generation and validation.
- AI training for human resources departments.
- Impact assessments and periodic audits to avoid discriminatory bias.
- Transparency and documentation in dismissal decisions.
- Internal policies on the responsible use of AI.
In conclusion, the article stresses that AI in labour management should be used with caution and under human supervision to avoid infringements of fundamental rights. The implementation of AI in labour must guarantee fairness and transparency, ensuring that efficiency is not achieved at the expense of workers' rights.
