Can ChatGPT recognize impoliteness?

Marta Andersson and Dan McIntyre are quoted in a recent article in the UK newspaper, The Guardian, concerning ChatGPT’s capacity to produce abusive language. The article also references Marta and Dan’s recent article in the Journal of Pragmatics, ‘Can ChatGPT recognize impoliteness?’ Marta and Dan are currently working with Meriem Beloucif from the Department of Linguistics and Philology to explore the impact of AI on language from a range of linguistic perspectives. Most recently this resulted in a one-day workshop entitled ‘Language in the Age of AI: Consciousness, Creativity, Philosophy, Ethics’, held on 19 March 2026. The event brought together 18 researchers from seven universities in Sweden, Germany and the UK. Work is currently ongoing to develop this network in order to further explore the issue of AI and language.