We share the article by lawyer Stefano Trifirò, published in Il Giorno Economia on May 19, 2025.
Robotics, algorithms, artificial intelligence, and machine learning are rapidly transforming the world of work. Stefano Trifirò, a lawyer at the firm founded by his father Salvatore Trifirò—one of the leading figures in Italian labor law—explains how the issue has been addressed in the courtroom.
“The first problems between AI and humans in the labor market arose years ago,” he says, “when job applications began to be processed through digital platforms, which managed hiring, oversaw the employment relationship, and could even terminate it. I’m referring to the work of delivery riders, that is, bicycle couriers.”
What happened?
“Some workers sued the company to have their employment classified as a subordinate work relationship. Among the various issues brought before the judge, it emerged that the artificial intelligence algorithm used to select riders not only discriminated based on political opinions, but also penalized them in their jobs. As early as 2020, the Bologna court sided with some of the workers, ruling that the algorithm—named Frank—violated constitutional rights at the foundation of our legal and labor systems.”
Were there earlier precedents?
“I remember a similar case, between 2014 and 2017, when Amazon developed an AI system to automate its hiring process. The algorithm was designed to analyze résumés and identify the most suitable candidates, based on past hiring patterns.”
What kind of programming errors occurred?
“In that case, the machine learning was managed by an all-male team, and the software learned that the ‘ideal’ candidate was, in fact, a man. The result? The software began to penalize female résumés, lowering the scores of applicants who had attended women’s colleges or had experience with women’s rights groups.”
Could similar cases still happen today, five years later?
“Since 2020, there have been major advancements. The AI Act—the new European regulation on artificial intelligence—classifies AI systems used in recruitment as ‘high risk.’ This means they must comply with strict rules on transparency, oversight, and reliability.”
