Avv. Alessandro Del Ninno took part as speaker in the Paradigma workshop “Supply Contracts for Artificial Intelligence Systems and Models”, held on 25 March 2026 and devoted to the contractual, regulatory and operational issues arising in the procurement and outsourcing of AI solutions. His presentation focused on the legal structuring of AI outsourcing arrangements, the allocation of roles and responsibilities across the supply chain, and the contractual safeguards needed in light of the AI Act and the evolving European framework.
During his intervention, Avv. Del Ninno examined the main contractual implications of the AI Act for suppliers, deployers and downstream operators, with particular attention to the distinction between high-risk AI systems, general-purpose AI models and other AI solutions. He highlighted the need for careful pre-contractual assessment of the system offered, especially where the solution may qualify as high-risk or may later be transformed into a high-risk use through customization or change of purpose.
A central part of the presentation was devoted to contract drafting methodology. Avv. Del Ninno explained that AI clauses should not merely restate legal obligations, but must translate regulatory and operational risks into executable, auditable and negotiable contractual provisions, supported by processes, evidence, service levels, remedies and technical annexes. In this perspective, he emphasized the importance of a coherent contractual architecture capable of governing data flows, compliance obligations, audit rights, liability, exit strategies and vendor lock-in risks.
He also addressed the safeguards that customers should require when acquiring high-risk AI systems, including documentary evidence on risk management, data governance, technical documentation, logging capabilities, human oversight, cybersecurity and quality management. Particular attention was devoted to the contractual governance of datasets, the use of personal and non-personal data, the correction of bias, and the need to verify that suppliers can demonstrate compliance across the full lifecycle of the system.
In the final part of his presentation, Avv. Del Ninno considered broader risk allocation issues, including outsourcing concentration risk, subcontracting chains, exit and portability clauses, and liability for damage caused by defective AI systems in light of Directive (EU) 2024/2853 on product liability. His contribution provided participants with a practical framework for negotiating AI supply contracts that combine regulatory compliance, operational control and effective protection against technological, legal and commercial risk
