When the Algorithm Decides: The Company's New Legal Liability

The growing integration of artificial intelligence systems into business processes is shifting the debate from technological efficiency to legal liability. In an article published in Expansión, María Pardo de Vera, a partner in the data protection and digital law practice, analyzes the challenges posed by automated decision-making and emphasizes that “AI is no longer just a technological or reputational issue; it has become a matter of legal liability.”

The expert points out that “the question is who is accountable when these systems are involved in decisions that affect people, rights, economic opportunities, or labor relations,” especially in areas such as hiring, granting funding, or handling complaints. In this context, she emphasizes the importance of organizations being able to explain “who made the decision, based on what information, according to what criteria, and what mechanisms were in place to correct potential errors.”

It also highlights that both the GDPR and the European Artificial Intelligence Regulation (AI Act) incorporate mechanisms designed to ensure traceability, human oversight, and the ability of companies to demonstrate how the systems they use operate and who is accountable for their results. As he warns, “a company that uses AI cannot hide behind the fact that the model was developed by a third party, that the result was generated automatically, or that the final decision was made by an employee who merely validated a recommendation.”

Finally, María Pardo de Vera warns that “the main risk does not always lie in the algorithm, but rather in the lack of internal governance,” and concludes that when a decision affecting a person is determined by an algorithm, “someone must be held accountable for it.”