The rapid adoption of generative artificial intelligence (AI) and neural machine translation (NMT) is reshaping professional translation practice and raising important questions about the evolving role of the linguist. This presentation explores how AI is transforming translation workflows and examines the continuing role of human expertise in producing high-quality translations. While public discourse often frames AI as potentially reducing the need for human translators, professional practice suggests a far more nuanced reality. Far from eliminating human involvement, AI redistributes linguistic expertise across the entire translation process, changing how linguists contribute to translation production.
Drawing on observations from the implementation of a structured AI-assisted translation workflow within a professional language services environment, the presentation explores how human expertise contributes at multiple stages of the translation process, including terminology management, prompt and technology selection, workflow design, quality assurance, revision, and final validation. Instead of treating AI output as an autonomous product requiring only correction, the proposed framework conceptualises translation as a collaborative process in which technological capabilities and human judgement operate in continuous interaction.
The presentation also briefly considers how the extent and type of linguistic intervention vary across different translation domains—technical, creative, legal, or institutional—shaped not simply by the quality of machine-generated output but by communicative purpose, audience expectations, regulatory constraints, and domain-specific conventions. This suggests that AI-assisted translation cannot be understood as a uniform process but must be considered in relation to the specific characteristics of different translation contexts.
These ideas are explored further in the accompanying workshop, From Translation Production to Translation Direction: A Practical Workshop for Linguists in the AI Era, which invites participants to put them into practice.
Beti Marina Pagourdaki is a Localization Project Manager and Language Lead in the Localization Department at ORCO S.A., where she brings together linguistic expertise, project coordination, client communication, and quality assurance. With eight years of experience across different roles in the language services field, she has worked extensively with legal, marketing, and web consumer-facing content—domains where “good enough” is rarely good enough, and where context, audience, and purpose make all the difference.
She holds a Degree in Law from the National and Kapodistrian University of Athens and has completed additional professional translation trainings. Her background gives her a particularly sharp eye for the practical decisions behind translation quality: terminology, tone, risk, workflow, and final validation. In the context of AI-assisted translation, Beti’s work reflects the modern translation industry’s central question: if AI can produce a draft, what does human expertise still contribute? Her experience positions her at the intersection of project strategy, linguistic judgement, and reliable delivery in a rapidly changing localization landscape.
