Italy has put its national machinery for enforcing the European Union's AI Act into operation. Legislative Decree No. 179/2026, dated October 7 and published in Italy's Official Journal on October 8, took effect on October 23, 2026. It assigns AI oversight, market surveillance, inspections, registration, and sanctioning powers across a set of national authorities — turning the EU's landmark AI law from a Brussels regulation into an enforceable reality in Europe's third-largest economy.
For startups selling AI products into Europe, the decree matters because it answers the question every founder asks about a new regulation: who exactly will come knocking, and with what powers?
Background: The EU AI Act and Its National Gap
The EU AI Act, Regulation (EU) 2024/1689, entered into force on August 1, 2024, and is the world's first comprehensive AI law. Its timeline has rolled out in stages: bans on prohibited practices and an AI literacy obligation applied from February 2, 2025; obligations for providers of general-purpose AI models applied from August 2, 2025; and most transparency obligations applied from August 2, 2026.
But the AI Act is a regulation that depends on member states to designate national market-surveillance authorities and set up enforcement structures. Until a country does that work, the law's obligations exist on paper with no domestic enforcer to apply them. In July 2026, the Digital Omnibus on AI, Regulation (EU) 2026/1744, adjusted the schedule further, postponing high-risk obligations to December 2, 2027 for stand-alone systems such as those used in recruitment, credit scoring, or education, and to August 2, 2028 for AI embedded in regulated products. It also added a ban on AI generating non-consensual intimate content or child sexual abuse material, applying from December 2, 2026.
The European Commission's AI Office has meanwhile begun exercising its own enforcement powers over general-purpose AI models, sending its first formal information requests to OpenAI and Anthropic in October 2026. Italy's decree is the national counterpart to that Brussels-level activity.
What Happened: Who Does What in Italy
Under Decree 179/2026, responsibilities are split rather than concentrated in a single AI regulator:
- AgID (the Agency for Digital Italy) assesses, designates, notifies, and monitors conformity-assessment bodies, including through periodic and extraordinary audits. These are the bodies that certify whether high-risk AI systems meet the Act's requirements.
- ACN (the National Cybersecurity Agency) holds general AI market-surveillance duties and may inspect real-world testing remotely or on site, without prior notice.
- Sector regulators — the Bank of Italy, CONSOB (the markets regulator), IVASS (the insurance supervisor), and Italy's data-protection authority — hold specific, bounded market-surveillance remits within their sectors.
- Sanctioning powers are assigned by authority and remit, rather than through one general fine amount.
The structure reflects a deliberate choice: AI touches every sector, so enforcement follows existing sectoral expertise instead of creating a monolithic AI agency. It also means a fintech startup and a health AI startup in Italy may answer to different regulators under the same law.
Why It Matters for Founders and Investors
First, the decree ends the grace period of ambiguity for companies operating in Italy. A startup deploying AI in the Italian market now faces identifiable authorities with inspection powers, including unannounced on-site inspections of real-world testing by ACN. Compliance documentation, risk assessments, and testing records are no longer theoretical deliverables — they are the materials an inspector can ask to see.
Second, the split structure raises the cost of regulatory mapping. A company operating across finance, insurance, and general consumer markets may need to track the expectations of several Italian authorities at once, plus the Commission's AI Office for general-purpose models. Investors evaluating European AI startups should ask whether the team has mapped its regulator set country by country, not just in Brussels.
Third, Italy's move sets a template other member states are likely to follow, and it sharpens the competitive picture. Startups that built compliance into their architecture early — data governance, model documentation, human oversight, incident logging — now hold an advantage over competitors scrambling to retrofit. For founders, the lesson is familiar but newly concrete: in regulated AI markets, compliance is a product feature, and in Italy it now has an address.
Angeled.com connects founders with investors.
GDPRGard, AI and the GDPR in 2026 guide, October 2026
Euronews / AI Office enforcement reporting, October 14, 2026
Read Next
Open Weights Are Not Open Business Models: Mistral Large 4 and the Licensing Trap
Courts Are Writing AI's Rules Now: Three Rulings Founders Should Read
Delhi High Court Orders 24-Hour Takedown of AI Deepfake Videos Targeting Chief Minister
EU Court Adviser Says AI Training Is Legal — but Gemini's Answers May Still Infringe Copyright
AI Regulation After the Midterms: What Founders Should Do Before the January Bills Arrive
DOJ Launches Super Intelligence Task Force to Police AI Misuse



