Stereotypes in Artificial Intelligence Systems

The increasing use of Large Language Models (LLMs) in a variety of fields has raised concerns about their potential to perpetuate stereotypes and generate biased content. Furthermore, GenAI systems are playing an increasingly important role in selection processes and candidate profiling. In this study, we examined how LLMs shape responses to Italian ungendered prompts, focusing on gender and professional bias. It is crucial to understand these risks in order to ensure that AI-based systems do not exacerbate social inequalities, but rather contribute to more equitable outcomes.