AI-Driven Credit Risk Models in B2B Trade Credit Insurance: Transparency and Accountability under the AI Act and Solvency II Directive

Sammanfattning

This legal research examines the regulation of AI-driven credit risk models in B2B trade credit insurance under the EU's Artificial Intelligence Act (AI Act) and the Solvency II Directive. The study analyses how these frameworks address algorithmic transparency, accountability, and fairness while maintaining market stability. The AI Act classifies credit assessment systems as high-risk AI, imposing strict requirements for documentation, human oversight, and bias mitigation. Solvency II complements this with prudential rules mandating robust governance of AI models. The research identifies potential regulatory overlaps and disproportionate burdens, particularly for smaller insurers. A comparative analysis of the U.S. approach reveals contrasting regulatory philosophies, with the EU's ex ante rules differing from America's more decentralized model. The study evaluates how these frameworks affect innovation and regulate AI models in trade credit insurance. Key findings suggest the current EU regime provides strong safeguards but may benefit from adjustments. Recommendations include tiered compliance requirements, enhanced regulatory sandboxes, and better coordination between financial and AI regulators. The research contributes original insights into the intersection of AI regulation and financial services, offering policymakers and industry stakeholders a framework to balance innovation with necessary safeguards in this rapidly evolving sector.