Italy’s Legislative Decree No. 179 of October 7, 2026 takes effect on Friday, October 23, and it does two things no other EU member state has yet done in one instrument: it requires every Italian employer to assess artificial intelligence inside its mandatory workplace risk file, and it reserves final decisions on hiring, dismissal, discipline and performance to a human being. The decree was published in the Gazzetta Ufficiale No. 234 on October 8 and runs to 54 articles.
The decree implements Italy’s national AI law, Law 132/2025, and aligns national law with the EU AI Act, Regulation (EU) 2024/1689. It takes effect with no transition period.
Article 42: AI Goes Into The Workplace Risk Document
The workplace-safety obligation sits in Article 42, which amends Italy’s Consolidated Safety Act, Legislative Decree 81/2008. That statute already requires every employer to maintain a Documento di Valutazione dei Rischi — the DVR, or risk assessment document — identifying workplace hazards and the measures in place against them. From October 23, an AI system that falls within the decree’s scope must be assessed inside that same document: the employer must record the risks the system poses, the prevention and protection measures in place, and the training and information given to affected workers.
The test for scope is functional, not a product list. An AI system is covered if it affects the organisation of work, production pace, how tasks are performed, or safety-relevant decisions. A scheduling platform, a workflow automation tool, an algorithmic task-routing engine or a production monitoring system can all qualify — including systems whose vendor does not market them as AI at all. The employer preparing the DVR applies the test in the first instance. Hazards the assessment is expected to weigh include technostress, anxiety about algorithmic monitoring and error sources from faulty model outputs.
Article 42 carries no penalty of its own, but it plugs into the existing DVR enforcement regime: under Decree 81/2008, failing to prepare or properly maintain a DVR exposes the responsible employer or manager to fines of roughly €3,557 to €9,113 or arrest of up to six months, and failures that contribute to a workplace injury can bring criminal prosecution. Enforcement sits with the National Labour Inspectorate, which already inspects DVR compliance across all sectors.
Article 41: A Human Must Make The Final Call On Jobs
The adjacent Article 41 reserves the final decision on the creation, modification or termination of an employment relationship — including disciplinary measures and performance evaluations — to a natural person exercising effective and autonomous decision-making power. A merely formal or nominal human sign-off does not meet the requirement, and a worker’s consent is not an exception. Decisions or dismissals taken in breach are void, and a worker may demand an intelligible explanation that identifies the role the AI system played and the main parameters it used.
There is a deliberate limit: on the reading reported consistently by Italian employment-law practitioners, searching for and screening candidates is not treated as a final decision, even where an automated system excludes a candidate from later rounds — so algorithmic CV screening remains lawful, with GDPR Article 22 still applying to fully automated decisions there. Remote Work Europe, which verified the decree, its Gazzetta reference and its entry into force against the primary record on Normattiva, notes the full text of Article 41 itself rests on that corroborated professional reporting.
Who Enforces Italy’s AI Rules
Beyond the workplace, Decree 179 settles who supervises AI in Italy. AgID, the Agency for Digital Italy, is the notifying authority that assesses and designates the bodies certifying AI systems. ACN, the National Cybersecurity Agency, is the market surveillance authority and the single point of contact with Brussels, while Banca d’Italia, CONSOB and IVASS cover high-risk systems in their sectors. The decree sets penalties of up to €35 million or 7% of worldwide turnover for the most serious prohibitions, mirroring the EU AI Act, and creates an Italian regulatory sandbox for AI.
A companion decree, Legislative Decree 160/2026, already in force since September 30, covers policing, biometrics and criminal liability, including an offence for failing to maintain required safety or human-oversight measures on high-risk AI. Together the two decrees complete the implementing framework that Law 132/2025 — the first comprehensive national AI statute in the EU — required within twelve months of its October 2025 entry into force.
Why It Matters
Decree 179 matters because it moves AI enforcement out of specialist technology regulation and into the paperwork every employer already files. The EU’s own high-risk employment obligations were deferred by the Digital Omnibus to December 2027 — the shift GEroundtable covered when the EU Council agreed its Omnibus position this week — but Italy’s route makes AI risk assessment a routine labour-inspection matter from Friday, years earlier, and it does so through a document, the DVR, that inspectors already audit. For vendors of scheduling, monitoring and workforce software selling into Italy, the functional test closes the “we are not an AI product” escape route: if the tool shapes how work is organised, it belongs in the customer’s risk file. The approach parallels the professional-responsibility line taken in Britain this week, where the Law Society told solicitors that AI is not a legal entity and responsibility cannot be transferred to it — in both countries, the accountability stays with the human institution, a principle GEroundtable’s EU AI Act explainer tracks across the bloc.
Background
Italy passed Law 132/2025 in 2025, becoming the first EU member state with a comprehensive national AI statute, and gave its government twelve months to issue implementing decrees. The EU AI Act itself entered into force in August 2024 with staged application: bans and AI-literacy duties from February 2025, general-purpose model rules from August 2025, and most remaining obligations from August 2026, with the high-risk employment duties deferred to December 2027 by the 2026 Omnibus. Italy’s two 2026 decrees — 160 on policing and criminal liability, 179 on authorities, training and work — are how Rome chose to run ahead of that timetable rather than wait for it.
Source: TechTimes, Remote Work Europe
0 Comments