GUIDA della Società

Legge dell'UE sull'intelligenza artificiale Monitoraggio post-commercializzazione e segnalazione di incidenti gravi

The EU AI Act requires providers of high-risk AI systems to operate a post-market monitoring system and report specified serious incidents.

Ascolta questa guidaNarrato dall'IA · ~3 min

In questa pagina3 minuti di lettura
  1. Panoramica
  2. Immersione profonda
  3. Implementazione nel mondo reale
  4. The Future of EU AI Act Post-Market Monitoring and Serious Incident Reporting
  5. Metterlo in pratica
  6. Continua a esplorare
  7. Domande frequenti

Panoramica

Providers and deployers have distinct duties, and serious-incident reporting is not the same as routine customer support or general model drift monitoring.

Immersione profonda

Article 72 requires providers to establish and document a post-market monitoring system proportionate to the nature of the AI technology and the risks of the high-risk system. It should actively and systematically collect, document, and analyse relevant data provided by deployers or collected through other sources about performance throughout the system’s lifetime. The system feeds updates to risk assessment and may lead to corrective action. It complements, rather than replaces, provider quality management, deployer monitoring, and other sectoral safety obligations. Article 73 covers reporting of serious incidents. Providers must report to the market surveillance authorities of the Member States where the incident occurred, subject to the Regulation’s timelines and process. The Act includes a special deadline for incidents involving death: report as soon as a causal relationship is established or suspected, and no later than ten days after awareness. Other incidents have different deadlines based on severity and awareness. Do not compress these into one rule; consult the operative article for the specific event. Deployers have duties to inform providers or relevant authorities when they identify risks or serious incidents in operation, following the channels required by the Act. They should preserve relevant logs and cooperate with investigations. A provider should establish usable reporting routes, triage evidence, protect personal data, and maintain a record of decisions. An internal ticket or user complaint may be an early signal but does not automatically meet the legal definition of a serious incident. The practical system links post-market evidence to action: detect a signal, assess whether it relates to the AI system, determine severity and reporting duties, notify the right authority within the applicable period, and take corrective measures when needed. Assign named responsibility and escalation coverage outside business hours. Make sure vendors and deployers can supply information quickly, while avoiding unsupported conclusions about causation before investigation.

Implementazione nel mondo reale

A provider detects a safety-relevant output pattern and checks the affected system versions and users.

A deployer preserves logs and notifies the provider after an AI-supported workflow contributes to a serious injury.

An incident lead records the first awareness timestamp and consults Article 73 for the applicable deadline.

A post-market review identifies a drift trend and documents corrective action before further harm occurs.

The Future of EU AI Act Post-Market Monitoring and Serious Incident Reporting

The European Commission’s current timeline makes most AI Act requirements applicable from 2 August 2026, with Annex III high-risk rules delayed to 2 December 2027 and product-embedded high-risk rules to 2 August 2028. Some requirements and duties already apply earlier or under other laws, so teams must check the system category and applicable date. As oversight capacity and standards develop, organizations should update incident playbooks and authority contacts rather than assume a single calendar applies to all systems. Keep dated records of the applicable text, guidance, and decisions so teams can explain their reasoning when rules or system purposes change.

Metterlo in pratica

Perché è importante

  • Rischio e sicurezza. I danni catastrofici e quotidiani dell’IA dipendono entrambi da chi comprende i rischi e da chi può agire.
  • Decisioni più chiare. L’alfabetizzazione pubblica e professionale determina la possibilità politica di una forte politica di sicurezza.
  • Tagliare il clamore. Spiegazioni chiare riducono la cattura da parte di montature pubblicitarie, PR di laboratorio e vaghi teatrini etici.

Attenzione

  • Trattare il rischio esistenziale come fantascienza mentre le capacità si aggravano.
  • Confondere la sicurezza del prodotto superficiale con l'allineamento in condizioni di elevata autonomia.
  • Lasciando il pubblico non inglese e non esperto solo con fonti di bassa qualità.

Prossimi passi

  1. Separare i rischi di danni al prodotto, uso improprio e perdita di controllo/disallineamento.
  2. Chiedi quali prove cambierebbero la tua opinione sulle tempistiche e sulla gravità.
  3. Preferire fonti primarie e valutazioni concrete alle affermazioni di marketing.
  4. Identifica un percorso d’azione: carriera, politica, finanziamenti o competenze, non solo consapevolezza.

Continua a esplorare

Newsletter gratuita

Ricevi il briefing quotidiano sull'IA

Tre storie verificate sull'IA ogni mattina nei giorni feriali, scritte in un inglese semplice. Gratuito per sempre, senza pubblicità.

Un'email ogni giorno feriale. Disiscriviti con un clic. Non vendiamo né condividiamo mai il tuo indirizzo.

Mettiti alla prova

Take the EU AI Act Post-Market Monitoring and Serious Incident Reporting quiz

Feedback immediato su ogni risposta e un certificato condivisibile con un ID verificabile una volta superato un corso.

Inizia il quiz

Sostieni l'istruzione gratuita sull'IA. AI Understanding è un'organizzazione non profit 501(c)(3) — nessun annuncio, nessun paywall, mai. Fai una donazione

Domande frequenti

What is EU AI Act Post-Market Monitoring and Serious Incident Reporting?

The EU AI Act requires providers of high-risk AI systems to operate a post-market monitoring system and report specified serious incidents. Providers and deployers have distinct duties, and serious-incident reporting is not the same as routine customer support or general model drift monitoring.

How does Article 72 support oversight after a high-risk system enters service?

The provider’s system collects and analyses lifetime performance information.

Where does Article 73 direct a provider to report a serious incident?

The article identifies national market surveillance authorities.

What special deadline applies to a death-related serious incident?

The Regulation sets a specific ten-day outer limit for death cases.

Does every complaint automatically qualify as a serious incident?

Complaints can trigger triage but are not automatically legally classified.