Reconfiguring Translation Expertise in the Age of Algorithmic Mediation: A Global Perspective on Agency, Knowledge, and Digital Resilience
Abstract
The accelerating integration of generative artificial intelligence into translation practices is reshaping the epistemological and professional foundations of the field. While large language models (LLMs) are often framed as tools that enhance efficiency and accessibility, their growing presence also signals a deeper transformation in how translational knowledge is produced, circulated, and legitimised across global contexts. Within this evolving landscape, translation expertise is increasingly mediated by algorithmic systems that both enable and constrain human agency.
This paper situates these developments within the broader dynamics of globalisation, where technological infrastructures operate transnationally and reconfigure disciplinary boundaries between education, research, and professional practice. Drawing on the notion of algorithmic culture and recent discussions on digital patronage, the study examines how translation experts are repositioned within socio-technical systems that subtly regulate visibility, authority, and decision-making processes. The paper approaches LLMs as structuring forces that reshape the conditions under which expertise is exercised and recognised.
The analysis focuses on three interconnected domains. In educational contexts, the increasing emphasis on technology-driven competencies reflects a global shift in curriculum design, yet it also raises concerns about the marginalisation of theoretical knowledge and the attenuation of cognitive engagement. In academic research, the reliance on externally developed tools risks limiting scholarly agency, positioning researchers as evaluators of systems rather than contributors to their conceptual design. In the professional sphere, translators’ roles are fragmented and their participation in decision-making processes becomes increasingly restricted as a result of platform-based workflows and market-driven dynamics.
In response to these developments, the paper proposes a theory-driven approach to digital resilience in Translation Studies. It argues that translation scholars and professionals can reposition themselves as active agents by engaging with AI technologies not merely as users but as co-designers of analytical tools. Drawing on an AI-based model configured by the researcher, the study demonstrates how HCAI (Human-Centered Artificial Intelligence) and EITL (Expert-in-the-loop) approaches can enable scholars to build their own systems grounded in theoretical frameworks. Furthermore, the paper highlights the continued importance of theoretical training by showing that individuals with a strong theoretical background are better equipped to critically interpret outputs generated by the model. Finally, from a professional perspective, the practical value of the theory-informed model is discussed through editorial feedback, illustrating its potential to support decision-making processes rather than replace them.
By foregrounding the interplay between theory, technology, and practice, the paper ultimately proposes a reconfiguration of the translation expert as a reflexive agent capable of navigating and shaping the evolving conditions of algorithmic culture.
