This paper presents a technical curriculum on language-oriented artificial intelligence (AI) in the language and translation (L&T) industry. The curriculum aims to foster domain-specific technical AI literacy among stakeholders in the fields of translation and specialised communication by exposing them to the conceptual and technical/algorithmic foundations of modern language-oriented AI in an accessible way. The core curriculum focuses on 1) vector embeddings, 2) the technical foundations of neural networks, 3) tokenization and 4) transformer neural networks. It is intended to help users develop computational thinking as well as algorithmic awareness and algorithmic agency, ultimately contributing to their digital resilience in AI-driven work environments. The didactic suitability of the curriculum was tested in an AI-focused MA course at the [Institute] at [University]. Results suggest the didactic effectiveness of the curriculum, but participant feedback indicates that it should be embedded into higher-level didactic scaffolding – e.g., in the form of lecturer support – in order to enable optimal learning conditions.

This paper presents the pedagogical integration of MTUOC, an open-source project developed at Universitat Oberta de Catalunya (UOC)—a distance-learning institution—to facilitate the training, fine-tuning, and integration of Neural Machine Translation (NMT) and Large Language Models (LLMs). The project consists of a modular suite of tools designed to streamline complex technical workflows for translation purposes. These components are currently utilised across research, industry knowledge transfer, and formal education. Specifically, the tools have been successfully implemented in a Bachelor’s degree in Translation and Interpreting and a Master’s degree in Translation Technologies. Furthermore, a pilot open course based on this framework received significant interest, reaching over 100 participants. This paper outlines the core components of the project, discusses the teaching experiences gathered in asynchronous environments, and describes the organisation of a forthcoming open course scheduled for October 2026. The results suggest that providing students with accessible, high-level interfaces for AI-based translation technologies enhances their technical autonomy and professional readiness.

In this paper we describe how we mimick the different steps in the historical development of NMT systems, all trained and evaluated on the same small data set. We do this for pedagogic reasons so students can see the effect of each of the steps on metrics like BLEU but also on qualitative examples, which the training scripts generate after each epoch. As MT paradigms, we discuss NMT training from scratch, finetuning pretrained encoder-decoder models, and finally prompt engineering for decoder only models. All models run in Kaggle sessions and all Python scripts and JuPyter notebooks are made available to the MT teaching community through Github and public Kaggle sessions.

As Large Language Models (LLMs) are increasingly integrated into professional translation, there is a growing need for pedagogical frameworks that move beyond simple trial-and-error prompting across any available tools used for translation tasks. This paper presents a framework developed in a university MA seminar on translation to foster responsible AI literacy as well as linguistically grounded prompt control and structured prompting. The approach also emphasises that AI is understood as a tool for suggesting possible translation solutions, while human translators remain the final decision makers. Integrating translation-oriented text analysis and corpus-informed feature extraction from parallel and compa-rable data, the approach teaches students to develop register-specific and model-interpretable instructions when translating specialised texts with the assistance of generative AI tools. During a one-semester course, students familiarised with prompting strategies and different tools, including commercial, freely available models, GDPR-compliant institutional infrastructure and local models. These steps and the comparative evaluation of these tools allowed the students to identify optimal configurations that conform best to professional quality standards for specialised texts, and they ultimately led to highly improved translation output compared to unsteered model results.

Drawing on 23 student projects from a fourth-year Machine Translation and Post-editing course, this paper examines how asking students to compare LLM and NMT outputs, interpret metric results, and justify a post-editing choice reveals their evaluative judgement. Students translated short specialised English Wikipedia texts into Catalan or Spanish, generated four system outputs, evaluated them using automatic metrics and human adequacy/fluency assessment, selected one output for post-editing, and justified their decision in written reports. The analysis combines descriptive counts from 23 projects with qualitative coding of the 22 cases sup-ported by written reports. Results show that students did not treat automatic metrics as final authority: final post-editing selections often diverged from metric rankings and were justified through adequacy, fluency, terminology, and expected post-editing effort. The study therefore does not compare systems under benchmark conditions; it analyses how students justified system choice within an au-thentic classroom assignment.

In this demo, we present COPECO-Speech, a multimodal post-editing workbench designed for translator training with AI-based translation technologies. It extends an existing pedagogical post-editing platform (COPECO) by integrating speech input and Large Language Model (LLM)-assisted editing. The workbench has different post-editing modalities and helps teachers annotate student tasks using either a shared or personalized annotation scheme. The system logs all interactions—including keystrokes, speech input, editing actions and LLM operations—enabling detailed analysis of post-editing processes.

With the rapid technologisation of translation, skills beyond post-editing (PE), such as data literacy, technology evaluation, and critical engagement with AI-based tools are becoming essential competencies for trainee translators. This paper presents a syllabus for a translation technology module that equips MA Translation students with end-to-end technology assessment skills, from engine selection and domain adaptation to automatic and human evaluation, post-editing, and reporting results. Large language models are integrated throughout, as translation engines, as a basis for prompting and domain adaptation strategies, and as a source of explainable quality estimation signals. Students apply these skills in a simulated client scenario as project-based learning. Trends on students’ engine selection and customisation preferences observed across three years suggest a gradual shift towards LLM-based tools, but traditional engines and built-in options remain a firm presence in their practices.

This paper reports on an exploratory, qualitative study of technology use in a Simulated Translation Bureau (STB) in a master’s programme in translation. The STB is a practice-oriented module in which students work on authentic translation projects and integrate a range of technologies, including AI-based tools, into their workflows. The paper addresses how students’ technology-related decision-making can be made visible and assessable, including students’ reflections on how they justify and verify the use of output-generating technologies, and how they describe perceived changes in their technology use over the course of the simulation. The study shows that the STB is an effective pedagogical model for teaching and assessing technology judgment in AI-enabled translation workflows, but also highlights the need to foreground broader ethical dimensions of AI literacy more explicitly in future iterations of the pedagogical design.

This paper proposes an approach to teaching data management in translation programmes, framing it as an extension of established documentation practices. It argues that data governance, sourcing, and processing are core competences for translators working with NMT and LLMs, and outlines a train-ing framework that supports responsible data reuse, quality optimisation, and professional agency.

As the adoption of large language models (LLMs) continues to redefine translation as a professional practice and the skills required to effectively leverage assistive technologies, the academic community is responding by proposing approaches to integrating generative AI into translator training (see Penet, Moorkens and Yamada 2026; Pym and Hao 2025; Kornacki and Pietrzak 2024). Such approaches are ideally situated within a field-specific AI literacy framework, such as that recently introduced by Krüger (2025) in extending on previous modelling of MT and data literacy for translation. Within the interaction dimension of his AI literacy framework, Krüger outlines a cognitive level, which, among other things, encompasses the translator’s awareness of how AI use can both augment and impair human intelligence (2025:18). From a metacognitive perspective, indicators of interaction-oriented AI literacy include the translator’s ability to recognize and articulate the affordances and constraints of AI systems in relation to their own performance, and to act accordingly. This paper presents a preliminary, process-oriented approach built around think-aloud protocols and screen recording that translation trainers and trainees can draw on to identify potential instances of augmentation (new-skilling) and impairment (skill-skipping, no-skilling, de-skilling) (Weßels and Maibaum 2026) in conjunction with the utilization of generative AI. In doing so, it contributes to the growing body of literature on critical AI literacy and informed pedagogical implementation of generative AI. References Kornacki, M., & Pietrzak, P. (2024). Hybrid Workflows in Translation: Integrating GenAI into Translator Training (1st ed.). Routledge. https://doi.org/10.4324/9781003521822 Krüger, R. (2025). Implementing generative artificial intelligence technologies in language industry workflows – A competence perspective. Lebende Sprachen, 70(1), 11–38. https://doi.org/10.1515/les-2025-0016 Penet, JC, Moorkens, Joss & Yamada, Masaru (eds.). 2026. Teaching translation in the age of generative AI: New paradigm, new learning?. (Translation and Multilingual Natural Language Processing 25). Berlin: Language Science Press. DOI: 10.5281/zenodo.17580856 Pym, A., & Hao, Y. (2024). How to Augment Language Skills: Generative AI and Machine Translation in Language Learning and Translator Training (1st ed.). Routledge. https://doi.org/10.4324/9781032648033 Weßels, D. & Maibaum, M. (2026). Vom Deskilling zum Newskilling. Forschung & Lehre, 2(2), 30-32. https://esv-elibrary.de/journal/article/99.160005/ful.2026.02.11

As an orchestration infrastructure, agentic AI systems now can plan and decompose the pre-defined goals into a sequence of steps, decide on external function calls, and coordinate one LLM or multiple LLMs with specialised roles. In this context, this position paper adopts a future studies “backcasting” ap-proach that starts with a desirable future, en-visioned as one in which AI-integrated trans-lation workflows are transparent, accountable, and aligned with human values; it then works backwards to examine how translator exper-tise should be reconceptualised to sustain meaningful human-in-the-loop participation. In this sense, the study first conceptualises the current translation-service provision as a sys-tem structured around managerial, mediation, and authorising roles. It then analyses how these roles may be changed and augmented within the agentic AI-orchestrated workflows. Building on the analysis, we propose a series of competences that should be cultivated to achieve the envisioned future: 1) evaluation grounded in advanced language competence; 2) situated and context-sensitive judgement informed by cultural and experiential knowledge; and 3) strategic procedural plan-ning in the design and oversight of agentic AI-orchestration workflows. The paper con-cludes with recommendations for future peda-gogical development and empirical research.