The Ministry of Labor unveiled an official guide on Friday that explains how incidence algorithms operate in everyday work life. The document, prepared by the department led by the second vice-president, Yolanda Díaz, emphasizes that societies are increasingly shaped by algorithms and that it is possible to steer math to benefit society, a point Díaz underscored.
Issued by the second vice-president, the guide represents the effort of a panel of experts chosen by the ministry. The team includes Gemma Galdón as coordinator and researchers from universities such as Anna Guines, Ana Belén Muñoz from UC3M, and Javier Sánchez-Handbag from the University of Córdoba, along with Adrián Todoli from the University of Valencia. Spanning about 30 pages, it seeks to define what an algorithm is, how it functions, and the type of information that companies are legally required to disclose to their employees.
The initiative, described as rider law, broadens the right of workers to obtain information from their employers. The standard requires firms to reveal their own algorithmic “black boxes” and to disclose the extent to which these systems influence decision-making. For example, the process by which staffing decisions are made at different stages of a project can involve an intermediate algorithmic recommendation. The guide raises questions about the parameters that shape these recommendations and why certain roles end up with more or fewer personnel. This is a matter that works councils may press for once the law is enacted.
Among the guide’s key messages is the importance of consistent human oversight at every stage of the decision process, not merely at a single point. The aim is to prevent bias or unintended discrimination throughout the entire chain of decisions.
One of the questions the algorithm guidance addresses asks whether a company uses automatic systems to profile individuals. Today, handling large data sets to select and filter profiles is a common use of such algorithms, particularly in procurement or hiring. Knowing whether a system exists and the criteria it uses to prioritize one person over another, or to exclude certain groups, is crucial to avoiding hidden or discriminatory bias. The guide also clarifies whether a tool is built on fixed code or adapts through continuous learning. The coordinator noted that this initiative intends to reduce information asymmetry between employees and employers.
Beyond protecting worker rights, the guide can serve as a reference for companies that buy or deploy third-party algorithms. It prompts inquiries about software changes and their purposes, asking what kinds of modifications have been made in a given installation and why. Two central questions recommended for organizations to consider when adopting such programs are whether updates were performed and what their impact might be.
The second vice-president indicated during the guide’s presentation that the deployment of an algorithm to monitor unpaid overtime would be introduced in the near future. The project known as MAX stands for More Algorithm for Less Overtime and is designed to curb excessive overtime and, in particular, unpaid overtime. In the country, millions of unpaid overtime hours have been reported, and existing regulations on time recording are in force to address this issue. This guide frames these practices as areas where algorithmic transparency and human oversight can help ensure fair work conditions.