Abstract
This paper examines how emerging AI-driven language technologies, particularly machine interpreting (MI), challenge the governance of public sector interpreting (PSI). Using Norway as a case study, the authors analyse the interaction between the Norwegian Interpreting Act and rapidly advancing speech translation technologies. They argue that the distinction between interpreter augmentation and automation is functional rather than technological, requiring risk-based governance focused on overuse, misuse, and underuse. Drawing on horizon-scanning research, market analysis, and stakeholder perspectives, the paper identifies growing pressures toward unregulated substitution of human interpreters. It concludes that MI cannot be prevented but should be governed through evidence-based, technology-neutral regulation, quality standards, risk assessment, and interpreter training