The Linguist

Chartered Institute of Linguists SPRING 2025 The Linguist 31 OPINION & COMMENT The communicative consequences Studying how AI affects human speech opens up new possibilities for exploring the development of language in the coming years, but it also raises concerns about its long-term effects. One significant issue is the potential loss of linguistic diversity. As AI systems rely on standardised language patterns, they tend to favour widely accepted forms of expression, which could undermine individuality and creativity. The use of AI-generated language may also make communication less spontaneous and authentic, removing the natural elements that make human conversation rich in empathy and emotion. Moreover, AI's influence could threaten the preservation of less commonly spoken dialects, regional variations and cultural expressions, which may not be sufficiently represented in AI training data. This could weaken cultural identity and lead to a more homogeneous global language ecosystem, less reflective of the diverse ways humans communicate across cultures. This extends to education, where AI-driven language could influence how English is taught and learnt by non-native speakers. As GenAI tends to favour robotic, set expressions, students may miss out on learning the cultural distinctions, regional variations and idiomatic phrases that are vital to mastering a language. These aspects are key to understanding a culture's values and communication style. If this trend continues, it could lead to a simplification of language learning, limiting not only how people speak but also how they connect with different cultures. Can we take action? To address these concerns, it is important to consider concrete strategies to preserve the richness of human language. For example, promoting diverse linguistic expressions and protecting regional dialects, as they carry cultural histories and reflect unique ways of thinking. Additionally, AI tools should be designed to respect human communication styles, rather than imposing rigid structures. By raising awareness about these issues and thoughtfully integrating AI into our communication, we can ensure that it complements, rather than diminishes, the complexity of natural human language. As we continue to integrate GenAI into our daily experiences, we must empower individuals to make more informed decisions when interacting with technology. It is important to remain vigilant as GenAI tools evolve, ensuring they enrich, rather than oversimplify, human speech. We should find a balance between technological advancement and the preservation of authentic human expression. After all, the goal is not to let machines define our language – or our identity – but to preserve the essence of what makes us humans, with technology as our ally. 7 Notes 1 Hohenstein, J et al (2023) 'Artificial Intelligence in Communication Impacts Language and Social Relationships'. In Scientific Reports, 13, 5487; https://doi.org/10.1038/s41598-023-30938-9 2 Gavilán, C (2024) 'Esos tics de lenguaje que nos delatan'. In En la raya; https://cutt.ly/ae6IG2wr 3 Gray, A (2024) 'ChatGPT "Contamination": Estimating the prevalence of LLMs in the scholarly literature'. In Computer Science; https://doi.org/10.48550/arXiv.2403.16887 4 Pérez Colomé, J (2024). 'Humans Are Already Repeating Words Learned from ChatGPT, such as "Delve" and "Meticulous"'. In El País (English version); https://tinyurl.com/yeyt9e8e 5 Ibid. 6 Giles, H, Edwards, AL and Walther, JB (2023) 'Communication Accommodation Theory: Past accomplishments, current trends, and future prospects'. In Language Sciences, 9, 101571 7 Article elaborated within the framework of the projects PID2022-141031NB-I00 and the Ibero- American Network of Research on Academic Integrity (Red-IA). Dr Cinta Gallent-Torres is a PhD Assistant Professor at the Universitat de València. Dr Rubén Comas-Forgas is a Full Professor of Applied Pedagogy and Educational Psychology at the Universitat de les Illes Balears. © SHUTTERSTOCK AI HUMAN MACHINE: An AI generated image

• Cover