Attorney Alessandro del Ninno participated as a speaker in Paradigma’s specialist workshop on the deployment of Artificial Intelligence systems and models in the insurance sector, with a particular focus on the applicable European and Italian regulatory framework and its practical implications for insurers, intermediaries and control functions.
His presentation provided a systematic analysis of the EU AI Act, covering its scope, the risk-based classification of AI systems, the regulation of general-purpose AI models and the allocation of obligations among providers, deployers, importers, distributors and other economic operators. Particular attention was devoted to prohibited AI practices, the requirements applicable to high-risk systems, transparency obligations for limited-risk systems and the applicable sanctions regime, together with the potential amendments and simplification measures contemplated under the Digital Omnibus initiative on Artificial Intelligence.
A substantial part of the session focused on organisations deploying AI systems in their own operations. Attorney Alessandro del Ninno examined the direct obligations imposed on deployers and the circumstances in which a deployer may assume the status and responsibilities of a provider, including where it makes a substantial modification to a system, changes its intended purpose or markets the system under its own name or trade mark. The workshop also addressed human oversight, log retention, the quality and representativeness of input data, ongoing system monitoring, suspension of use where anomalies arise, and the duty to provide appropriate information to affected individuals.
Specific attention was given to high-risk AI systems used in the insurance industry, particularly those employed for risk assessment and pricing in life and health insurance. The workshop analysed the obligation to carry out a Fundamental Rights Impact Assessment, its required content and its relationship with the Data Protection Impact Assessment under the GDPR. The discussion covered risks of algorithmic discrimination, lack of transparency, excessive reliance on automated outputs, poor data quality and insufficiently representative datasets, as well as the governance measures required to ensure meaningful human oversight and responsible system management.
The presentation further examined the broader implementation ecosystem surrounding the AI Act, including European Commission guidelines, the AI Act Service Desk, compliance tools, the Code of Practice for general-purpose AI models, model contractual clauses for the procurement of AI systems, and the ongoing CEN and CENELEC standardisation work on AI risk management and quality management systems.
The final part of the workshop focused on Italian Law No. 132/2025, addressing the provisions already in force, the legislative delegations granted to the Government, the designation of national competent authorities, and the rules governing professional services, copyright, liability, user protection and criminal sanctions. Through a legally rigorous and highly practical approach, the session provided participants with a comprehensive framework for integrating Artificial Intelligence into insurance operations while ensuring regulatory compliance, effective accountability and robust governance of AI-related risks.
