Italy

AI behavior law: bot and agent disclosure, crawler and training-data rules, automated-agent transactions, and algorithmic decision-making.

0
Instruments
0
Enacted
0
Proposed / in discussion
high
Confidence

Summary

Italy presents medium risk for commercial web scraping. Codice penale art. 615-ter criminalises unlawful access to a computer or telematic system protected by security measures, or remaining within it against the will of the person entitled to exclude access. Ignoring ToS anti-scraping provisions or bypassing technical protections can constitute art. 615-ter criminal liability. Law 28 June 2024 n. 90 (Cybersecurity Act) tightened the art. 615-ter framework. Italy transposed DSM Directive 2019/790 via Decreto Legislativo 8 November 2021 n. 177, adding arts. 70-ter (research TDM, mandatory exception) and 70-quater (commercial TDM, subject to machine-readable opt-out) to the Legge sul Diritto d'Autore (LDA). Commercial TDM is lawful where the rightsholder has not set a machine-readable reservation. The Garante per la protezione dei dati personali issued comprehensive scraping guidance in September 2023 and, in May 2024 (decision n. 329), issued a dedicated guidance document on web scraping for generative AI training recommending counter-measures including robots.txt intervention. In December 2024 the Garante fined OpenAI €15 million — the first GDPR generative-AI fine in Europe (later annulled by the Court of Rome in March 2026 on procedural grounds). Italy retains the full EU sui generis database right (LDA arts. 102-bis to 102-quater). GDPR applies through robust Garante enforcement. No scraping-specific statute exists beyond the general criminal and IP framework.

Automated-access legality

Carried forward from the crawler-law index. Governs whether automated clients may access public websites in this jurisdiction.

DimensionValue
Authorization testsecurity mechanism bypass
Public-page carve-outunsettled
Terms-of-service browsewrap enforceablenotice dependent
Terms-of-service clickwrap enforceableyes
Copyright exception modeltdm dual track
Text and data mining — commercial statuswith optout
Text and data mining — opt-out mechanismrobots txt
robots.txt legal weightevidentiary
AI training-specific lawbinding
Privacy regimeGDPR
Trespass to chattelsnot recognized

Last reviewed: 2026-05-23. Confidence: high. Fast-moving area — verify before relying. Not legal advice.