A newly published reflection paper examines the reasons why generative AI should not be relied upon for translation and interpreting, particularly in sensitive contexts. Contrary to common claims made by major tech companies, AI is far from being a reliable gateway to multilingual communication.
Summary
This reflection paper explores the use and impact of Large Language Models (LLMs) in multi- and interlingual communication and mediation. It aims to raise awareness among users, both non-experts and professionals, as well as policymakers, regarding:
- how LLMs “communicate,” how their processes differ from human communication, and the implications of these differences when generative AI is used for translation and interpreting (Section 1);
- the current and future impact of AI on translation and interpreting professions, including communication quality, workflows, job tasks, and job satisfaction (Section 2);
- potential unintended side effects that must be carefully considered alongside AI’s benefits (Section 3);
- whether and how AI should be integrated into translation and interpreting education and training, both in formal academic programs for future professionals and in continuing professional development (Section 4).
You are invited to consult the full paper attached.