Overview

• Parliament approved the measure on Sept. 17, aligning the framework with the EU AI Act's risk-based approach.
• Cross-sector rules mandate traceability and human oversight in healthcare, education, justice, public administration and sport, with worker notification when AI is used on the job.
• New offenses criminalize unlawful dissemination of harmful AI-generated content, such as deepfakes, with penalties of one to five years in prison, and increase sentences for AI-enabled fraud and identity theft.
• Children under 14 can access AI services only with parental consent, and in healthcare patients must be informed when AI assists care while doctors retain final decisions.
• The law designates the Agency for Digital Italy and the National Cybersecurity Agency as lead authorities, preserves powers of sectoral watchdogs, defines copyright and text-and-data-mining limits, and authorizes up to €1 billion in state-backed investment that critics deem modest.