1st Workshop on Artificial Cultural Reasoning

Computerlinguistik

This workshop, investigates whether current AI systems are capable of cultural reasoning, like handling conflicting preferences due to varying cultural backgrounds of users. Are such capabilities required and (how) can they be machine-learned?

 

It will be held at Trier University and is part of the CURIA-GR project, awarded by the University of the Greater Region as part of the Guest Professorship scheme.

AI and Language Models Across Borders: Whose Culture Do They Represent?

While AI systems are often presented as globally applicable, their training data and design choices are shaped by particular linguistic and cultural contexts. The multilingual Greater Region of SaarLorLux, with its overlapping traditions and language across Luxembourg, Germany, France, and Belgium, provides an ideal setting to examine whether AI models can recognise differences between closely related cultures or whether they tend to reproduce dominant perspectives at the expense of local or minority ones.

Participants will discuss questions of evaluation, fairness, and responsibility: How can cultural sensitivity in AI be meaningfully assessed? Should language models adapt to local contexts or remain culturally “neutral”? And what challenges emerge when AI misrepresents or overlooks cultural distinctions in border regions where identity is highly contextual? The workshop will feature one or two keynote talks, structured discussions, and a poster session for which abstracts will be invited. By focusing on the cross-border dynamics of the Greater Region, the event aims to create a space for collective reflection on how cultural reasoning in AI can be better understood, tested, and improved.

Organisers

Dr. Alistair Plum (Université du Luxembourg)
Prof. Dr. Achim Rettinger (Universität Trier)

Abstract Submission

We invite the submission of abstracts that broadly deal with the theme and questions (see below) of the workshop. Accepted abstracts will be presented at a poster session during the workshop.

  • To what extent can LLMs distinguish between related but distinct cultures (e.g. Luxembourgish vs. German vs. French)?
  • How do we measure “cultural sensitivity” in a way that is both rigorous and fair?
  • Should AI be culturally adaptive (changing responses based on user’s context), or culturally neutral?
  • What risks emerge when AI fails to respect local cultural norms in multilingual, border regions like the Greater Region?
  • How can LLMs be trained to improve their cultural reasoning skills?

-> Send abstracts (300 words) to alistair.plum@uni.lu

Please note that abstract submission is optional, and that abstracts will not be published as part of the workshop.

Important Information

Dates
Workshop: 17/11/25 10:00 - 17:00
Registration Deadline: 13/11/25
Abstracts due: 31/10/25
Poster notification: 05/11/25

Location
Room K101 (old chapel)
Campus II
Trier University

Registration
Anyone interested in participating is invited to register (free) at the link below.
https://forms.gle/RHC8iREGQrKQZ5Xp8

Contact
alistair.plumunilu

Further information
https://curia-gr.github.io/