Transfer pricing, tariff-induced risk, and artificial intelligence supervision: a networked compliance model for banks and regulators

This article develops a simulation-based framework linking tariff shocks, transfer pricing (TP) distortions, AI-supported compliance capacity, and the network structure of multinational enterprises. Stylised simulations indicate three consistent patterns: tariff exposure is associated with greater deviation from arm’s-length pricing and wider profit misallocation; stronger AI compliance capacity reduces these distortions; and mitigation effects are stronger when compliance capabilities are deployed at structurally central entities within the network. The framework has indirect prudential relevance. Indicators derived from tariff exposure, pricing deviation, and network centrality may serve as informational overlays for monitoring large cross-border corporate borrowers, helping identify cases where intra-group opacity warrants closer supervisory or credit risk review. The study is exploratory: results derive from stylised simulations rather than observed firm-level data and should be interpreted as indicative patterns within an assumed system.