Source: La Capitale

The global regulatory landscape for artificial intelligence is accelerating but fragmented, with Western jurisdictions prioritising citizen protection, transparency and competition, and Eastern models tending toward state-centred, vertical management. The European Union adopted a comprehensive risk-based framework through the EU AI Act (Regulation UE 2024/1689), which entered into force in August 2024 and is directly applicable across member states; absolute prohibitions (including social-credit systems and mass biometric profiling) took effect in February 2025, while general transparency obligations are scheduled from August 2026. In May 2026 EU institutions approved a “Digital Omnibus on AI” to postpone some high-risk system deadlines—notably in critical infrastructure and healthcare—shifting certain compliance dates to late 2027–2028 to give businesses more time without altering the regulation’s core.

Italy enacted a national framework ahead of some EU partners with law 23 September 2025, n. 132, embedding anthropocentrism and requiring final decisions in critical sectors (health, justice, public administration, work) to remain with a human. February 2026 legislative proposals added rules for AI in social services and imposed a three-part obligation when algorithms evaluate citizen applications: clear prior notice at the procedure outset, traceability of decision criteria, and a right to human review (human-in-the-loop) allowing contestation and re-examination by a public official. China favours targeted, sectoral rules administered by central authorities, including anti-deepfake measures, watermarking, real-identity verification and registration/review obligations for large language models. The United States lacks an overarching federal AI law, relying on executive orders and agency actions plus varied state statutes (e.g., Colorado’s AI Act from 2026), while the UK uses a sector-led regulatory approach; Peru has an early national AI law focused on transparency and oversight, and Japan relies mainly on flexible guidelines with progressively stricter compliance in procurement and audits.