Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

Diritto al Digitale

Diritto al Digitale artwork

Italy’s New AI Law: What Changes for Corporate Liability?

Oct 29, 20266 minAudio
ShareXLinkedInFacebookWhatsAppEmail

How does Italy’s new Artificial Intelligence Law change the liability risks for businesses, and what should companies do to protect themselves? In this episode of The Legal Break Podcast, Giulio Coraggio, technology and data lawyer, and Antonio Ravenna, journalist, discuss the impact of Italy’s new AI legislation on corporate liability. They explore how the new framework affects companies developing, deploying or using AI systems, the potential exposure of businesses and their management, and the practical steps organisations should consider to mitigate legal risks. From AI governance and accountability to risk management and compliance, this conversation examines what the new law means for businesses and how companies can prepare for the evolving legal landscape surrounding artificial intelligence. Listen to the episode to understand the implications of Italy’s new AI law for corporate liability and the key issues businesses should address when adopting AI. Send us Fan Mail

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

SPEAKER_01Hello everyone, I'm Giulio Coraggio, a technology and a data lawyer. And I'm Antonio Ravenna, a journalist. Today we're gonna discuss about the new law that introduced in Italy a specific liability regime for damages caused by artificial intelligence systems.

SPEAKER_00And we're gonna make it through the time of a coffee break. So, Giulio, the Italian law introduced a new liability for AI. Can you guide us through it?

SPEAKER_01Yes, uh it's a peculiarity of Italy, it is really connected to the underlying uh regime that we have in the country. The legislator introduced uh two crimes uh relating to violations and damages caused by artificial intelligence systems, so connected to the lack of uh security measures that cause a damage to the affected individuals. These crimes uh lead to a personal liability of whoever, for instance, in an organization, is liable for the violations, but that's not all.

SPEAKER_01On top of that, the regime provides for a corporate criminal liability. So on top of the liability of uh the managing director, of uh the chief uh compliance officer, of uh other individuals that um could be considered as um the main stakeholders within the organization for AI compliance and security, then there could be a corporate criminal liability of the company itself that leads to financial fines and uh to the disqualification from uh public tenders, determination of licenses, which can be very impactful for the company.

SPEAKER_01In addition to that, there is also a civil liability regime. When damages are caused by artificial intelligence systems, there is a reversal of the burden of proof. So if the nexus is proven, then uh it shall be on the company to prove that the damages were not caused by the artificial intelligence system. And in addition to that, there is the possibility by the affected individuals to claim that the insurance company covering the liability of the company towards those kind of claims can pay directly the affected individuals for these damages, which uh will lead to potential additional uh uh recalculations of the insured damages by insurance companies because they were got by surprise uh due to this regulatory change.

SPEAKER_00So it's a full package, we have civil, criminal, and administrative liabilities now.

SPEAKER_01Yeah, exactly, and it's everything peculiar of Italy. This regime goes on top of the sanctions provided by the AI Act or other regimes that can be affected by uh AI systems.

SPEAKER_00And so what companies should be doing now?

SPEAKER_01Well, on the corporate criminal liability regime, Italy has uh a specific legislation, it's commonly called the 231 regime. Companies need to have uh in place a corporate criminal compliance system in place, which is a protocol made of specific procedures that need to be tailored on the Italian regime. So when uh clients tell me we have an ethical code, no, that's not enough. It needs to be tailored on the specifics of Italian law, and that is crucial because it allows to exempt the company from the uh liability that will affect the individuals.

SPEAKER_01So basically, you will have a criminal sanction against the individuals within the organization that are deemed to be liable for the crime, but that would not cause a domino effect against the company. There wouldn't be an additional liability of the company uh on top of uh the liability of the individuals. In relation to the civil liability regime, the solution is not easy. You need to monitor what your company does with artificial intelligence systems, mapping the AI systems that your employees are using, the purposes, the use cases that they adopt is crucial.

SPEAKER_01It's not only important to have in place an approval process at the time you contract an AI system, you need to monitor how they use it because the regime is based on the purpose, the level of risk provided by the AI Act Queen, and so the uh risk-based assessment is triggered by the purpose for which the AI system is used rather than just the onboarding of that technology. It is burdensome, it is difficult to map, but it's crucial for any sort of organization that should have a uh compliance framework in order to monitor the risk.

SPEAKER_01So map and control everything. Well, everything uh seems burdensome, but in a system, in a solution that is affecting basically every organization. I always say that um AI, unlike other technical solutions or AI or other regimes, cuts across the whole organization. That's true because AI will impact any or is already impacting any department of an organization, and you need to really have a sort of spider that goes with through your organization to monitor what everyone is doing and uh make in uh put in place the uh adequate safeguards to avoid that uh misbehaviors are performed.

SPEAKER_01Thank you, Julia. Thank you.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. How to deal with a vulnerability and an incident under the Cyber Resilience Act?

    Oct 8, 20266 minAudio

    What should a company do when it discovers a vulnerability in a product covered by the EU Cyber Resilience Act? And what happens when that vulnerability becomes a cybersecurity incident? In this episode, Giulio...

    Credit the people on this episode

  2. Cyber Resilience Act: What Incidents and Vulnerabilities Must Be Reported?

    Oct 1, 20265 minAudio

    What incidents and vulnerabilities must be reported under the Cyber Resilience Act (CRA)? Since 11 September 2026, the CRA reporting obligations are in force, with companies required to assess and report certain...

    Credit the people on this episode

  3. Cyber Resilience Act: What Does the September 11 Deadline Mean?

    Sep 30, 20265 minAudio

    The September 11 deadline under the Cyber Resilience Act has passed, and the reporting platform is already operational. Companies within the CRA’s scope need to understand what the new reporting obligations mean and...

    Credit the people on this episode

  4. Prediction Markets in Europe: Are They Gambling?

    Sep 16, 20266 minAudio

    Prediction markets are expanding rapidly, allowing users to trade contracts linked to the outcome of future events, from elections and sports competitions to economic and other real-world events. But how should these...

    Credit the people on this episode

  5. EDPB Anonymisation Guidelines: What They Mean for AI Systems & GDPR

    Sep 4, 20268 minAudio

    The EDPB has issued its new Guidelines on Anonymisation, and they could have major implications for AI systems, AI training data and GDPR compliance. In this episode of the Legal Break, Giulio Coraggio discusses the new...

    Credit the people on this episode