Italy Bars AI From the Final Call on Hiring and Dismissals
Legislative Decree 179/2026 entered into force on 23 October 2026. Under Article 41, recruitment, contract changes, dismissals, discipline and performance assessments cannot be decided by automated systems alone, and a named person has to hold the decision.

Italy's Legislative Decree No. 179 of 7 October 2026 came into force on 23 October, and with it a rule that no algorithm in an Italian workplace gets the last word. Under Article 41, recruitment, contractual changes, dismissals, disciplinary measures and performance assessments cannot be made exclusively by automated systems. The final decision has to be taken by a natural person holding effective and independent decision-making authority.
The decree was published in Gazzetta Ufficiale, Serie Generale No. 234, on 8 October 2026, which records the entry into force as 23 October. It completes the delegation the Italian parliament handed the government in Article 24 of Law No. 132 of 23 September 2025, the national AI law, which sits alongside the EU AI Act rather than replacing it.
What Article 41 asks of a manager
DLA Piper's employment practice, writing on 21 October 2026, reads Article 41 as covering the full set of decisions that change someone's job: hiring, changes to the contract, dismissal, discipline and performance review. Employees are also entitled to clear and understandable information about which AI systems are in use and how far those systems shape the outcome. The requirement is not that AI stays out of the process. It is that a person holds the decision and can be shown to have held it.
Article 42 does something quieter and probably harder. Where AI systems affect how work is organised or how production runs, employers have to fold them into the workplace risk assessment that Italian law already requires of every company, and provide workers with information and training on the risks those systems carry. Articles 47 and 48 attach AI literacy and training duties for professionals and for employment services staff.
The ground this lands on is not empty. A European Commission survey published on 14 October 2026, which Parallax Nexus reported, found that 42 per cent of EU small firms and mid-caps already use AI or algorithmic management, mostly experimentally, while only 2 per cent of their decisions are fully automated. Italy's decree is written for the gap between those two numbers, which is the large and growing set of decisions where software ranks, scores or recommends, and a person signs.
The price of getting it wrong
A&O Shearman's analysis of the decree describes a graduated penalty scheme. It runs from one million euros, or 0.5 per cent of total worldwide annual turnover where that is higher, for failure to meet the deployer obligations in Articles 27 and 86 of the EU AI Act, up to 7.5 million euros or 1 per cent of worldwide turnover for supplying incorrect, incomplete or misleading information to the competent authorities. The AI Act's own ceiling for prohibited practices, 35 million euros or 7 per cent of worldwide turnover, sits above all of it.
Supervision is split several ways. The law firm Legance describes the decree as completing the supervisory, inspection and sanctioning toolkit available to AgID, the national cybersecurity agency ACN, the data protection authority and the financial sector regulators, with the Ministry of Labour coordinating measures on health and safety at work. Decree No. 160 of 9 September 2026, the companion instrument, carries the criminal provisions.
Easy on paper, hard in a tribunal
The honest reading is that Article 41 is simple to comply with on paper and difficult to comply with in fact. What is not known is how an Italian labour inspector or court will test effective and independent decision-making authority. No case has yet asked whether a manager who approves a software-ranked shortlist in under a minute, on a screen showing a score and a recommendation, is exercising judgment or ratifying someone else's. Until one does, the rule's reach is a question about evidence rather than about drafting.
The EU AI Act's deployer transparency obligations took effect on 2 August 2026, with the high-risk duties partly postponed, according to DLA Piper's tracking of the regime. Italy has not waited for the rest of it. Employers with Italian staff have had the text since 8 October and the obligation since 23 October.
What happens next?
- Italian employers using AI in recruitment or performance management must document who holds the final decision and inform staff which systems are in use.
- AgID, ACN, the data protection authority and the financial sector regulators now hold inspection and sanctioning powers under the decree, with no enforcement case yet reported.
- Other EU member states are still legislating their national AI Act implementations, and Italy's wording on human authority over employment decisions gives them a drafting precedent.
Related topics
Sources & references
- 01Decreto Legislativo 7 ottobre 2026, n. 179 — Gazzetta Ufficiale della Repubblica ItalianaprimaryOfficial Gazette entry, Serie Generale n.234 of 08-10-2026, recording entry into force on 23 October 2026.
- 02Italy introduces workplace rules on artificial intelligence — DLA PiperreportPublished 21 October 2026. Source of the reading of Articles 41, 42, 47 and 48.
- 03Italy implements the AI Act: Key legal requirements — A&O ShearmanreportAccessed 26 October 2026. Source of the graduated penalty figures and the link to Decree No. 160 of 9 September 2026.
- 04Italy's AI Rulebook Takes Shape: another decree on AI supervisory authorities, workplace safeguards and trade-secret protection — LegancereportAccessed 26 October 2026. Source of the division of supervisory responsibility and the Ministry of Labour's coordinating role.
More from Future
Barclays Puts Developers at the Centre of 5,000 Job Cuts
Barclays will cut around 5,000 full-time roles, about 5 per cent of its workforce, by the end of 2026. Anna Cross tied roughly 60 per cent of the actions to support functions, including fewer developers, in one of the first such statements from a large bank.
Anthropic's Junior Technical Hiring Fell From 27% to 9%
Anthropic's own employment data, published on 16 October 2026, shows the junior share of technical hires falling from 27 per cent to 9 per cent since early 2025, while agent hours outnumbered employee hours six to one by August 2026.
Small Firms Expect AI to Add Jobs While Sales Stay Flat
Will Aarons and Asani Sarkar used the 2025 Small Business Credit Survey to compare the twelve-month expectations of AI adopters and non-adopters among 5,248 US employer firms. The gap survives controls for firm age, size, profitability, owner characteristics and past employment, and is widest among firms in poor financial condition.
Danish Records Show No Pay Effect From Two Years of AI Chatbots
Anders Humlum and Emilie Vestergaard link Statistics Denmark surveys to administrative earnings records and find no effect of AI chatbots on pay or hours across 11 highly exposed occupations. New tasks, employer tooling and occupational switching all moved instead.



