Artificial intelligence has now become an integral part of business processes: recruitment, performance evaluation, work organization, training, and document management are just some of the areas in which AI tools are increasingly being used.
While companies are experimenting with new applications and seeking to reap their benefits in terms of efficiency and productivity, the legislator is defining rules and responsibilities. In this context, the first drafts of the implementing decrees of the Italian delegated law on artificial intelligence, recently approved on a preliminary basis by the Government, come into play.
These texts are therefore not yet final and may be subject to changes during the institutional process; however, they already clearly indicate the direction taken by the regulator and introduce elements that are destined to have a concrete impact on HR functions, compliance, and corporate governance.
No fully automated decisions affecting employees
One of the most significant principles concerns the prohibition on relying exclusively on automated systems for decisions affecting the employment relationship.
Hiring, employment management, disciplinary measures and terminations may not be the result of a decision made solely by an algorithm. Final responsibility must remain with a natural person vested with effective autonomous decision-making power. In other words, human oversight may not be merely formal but must be substantive and, where necessary, demonstrable.
This is a particularly relevant development for all organisations that are introducing AI systems into HR processes, particularly where such tools produce classifications, scoring, recommendations or assessments capable of influencing managerial decisions.
Continue reading the full version published on Ai4Business