(DE)GENERATIVE AI AND MULTILINGUAL (MIS)COMMUNICATION
Kris Peeters (University of Antwerp)
20 April, 5:30 PM
On April 20 at 5:30 PM, we had the pleasure of hosting an engaging online seminar by Kris Peeters (University of Antwerp), titled (De)Generative AI and Multilingual (Mis)Communication. The talk addressed one of the most pressing and controversial developments in contemporary language use: the rapid rise of Generative Artificial Intelligence (GenAI) and its impact on multilingual communication.
Since the emergence of tools such as ChatGPT, Gemini, Copilot, and Claude, the use of Large Language Models (LLMs) has expanded dramatically across sectors, including business, media, legal services, education, and public administration. While these technologies promise unprecedented accessibility to multilingual communication, they also raise critical concerns about the outsourcing of human thinking and the erosion of linguistic expertise.
The seminar explored these tensions from a critical perspective. After introducing the Special Interest Group on AI in Translation and Interpreting—established within the European Language Council—Peeters discussed key insights from the group’s recent reflection paper, AI for Translation and Interpreting: A Roadmap for Users and Policy Makers. This publication examines how LLMs “communicate” compared to humans, the transformative effects on the language professions, and the broader legal and ethical implications of adopting GenAI technologies.
Particular attention was devoted to the impact on translators and interpreters, whose roles are increasingly reshaped by automation, as well as to the risks associated with overreliance on AI-driven communication tools. The talk also addressed the place of GenAI in higher education, questioning how—and whether—it should be integrated into academic curricula.
Throughout the seminar, complex concepts such as embeddings, deep learning, bias, hallucination, and stochastic parroting were introduced and critically unpacked, offering participants a clearer understanding of both the potential and the limitations of current AI systems.
The event sparked thoughtful discussion among participants and highlighted the urgent need for informed, critical engagement with AI technologies in multilingual contexts.
A recording of the seminar is available at the following link: 2026 CELELC online seminar (De)generative AI and Multilingual (Mis)communication – prof. Kris Peeters-20260420_174141-Registrazione della riunione.mp4