Evaluating the Acceptance of Computer-Assisted Interpreting Tools Using the Technology Acceptance Model

Abstract

The adoption of Computer-Assisted Interpreting (CAI) tools among professional interpreters remains relatively limited. Factors such as high cognitive load, time pressure, fragmented workflows, and perceived usability challenges contribute to resistance in professional interpreting settings. This study addresses this gap by proposing the development of an Artificial Intelligence (AI)–enhanced CAI tool, designed through a human-centered methodology and evaluated within a framework based on the Technology Acceptance Model (TAM). Beyond a purely technological approach, the research adopts an ergonomic perspective aimed at understanding how interpreters perceive, adopt, and integrate technological tools into their professional practice. Semi-structured interviews were conducted with interpreters from diverse professional backgrounds, experience levels, and age groups. Guided by the TAM constructs — Perceived Usefulness (PU), Perceived Ease of Use (PEOU), and Behavioural Intention to use (BI) — the study explores participants’ technological competence, current professional needs, expectations regarding CAI tools, and perceived shortcomings of existing solutions. The interviews indicate that, although interpreters generally view technology as a potentially valuable ally, current tools are perceived as lacking in accuracy, reliability, and cognitive ergonomics. A widespread lack of formal training in technological tools further affects confidence and willingness to adopt new systems. Participants emphasized the need for context-sensitive, high-precision solutions capable of real-time transcription, advanced speech recognition, and automatic glossary generation, while consistently underscoring that artificial intelligence should function as support rather than replacement. Finally, an integrated CAI tool is proposed that combines preparation support, in-booth assistance, and pre-task training within a unified environment. Core functionalities include glossary extraction based on Natural Language Processing (NLP), real-time transcription and translation powered by Automatic Speech Recognition (ASR) and Neural Machine Translation (NMT), contextual terminology highlighting, and realistic training simulations using speech technologies.

Keywords: Computer-assisted Interpretation, Technology Acceptance Model, Automatic Speech Recognition, Text-to-speech, Interpreter Interview

DOI: 10.54941/ahfe1007296