5th Luxembourg Workshop on AI and Epistemology

The general goal of this workshop is to explore philosophical issues lying at the intersection of AI and epistemology. In our experience, such issues typically do not stay within the borders of epistemology, but also touch on themes from, for example, the philosophy of science and the philosophy of mind. Accordingly, the thematic scope of the workshop is broad. Thus, questions of interest include, but are certainly not limited to, the following:

(i) How (if at all) is it possible to understand, explain and gain knowledge about black-box AI systems given the complexity and opacity of their internal operations and training history?

(ii) How might AI technologies be used to supplement and improve our own human epistemic capacities?

(iii) When (if ever) is it rational to rely on AI technologies whose internal operations we do not fully understand when forming beliefs?

(iv) What fixes the content of the outputs of Neural Networks? When (if ever) should we attribute contents to internal parts/processes of Neural Networks?

(v) To what extent, and in what ways, are the linguistic outputs of Large Language Models similar or dissimilar to Human Testimony?

This workshop is part of the FNR funded project ‘The Epistemology of AI Systems’ (EAI) which is wrapping up in 2026. It is also the 5th in a series of workshops on Artificial Intelligence and epistemology. Three of these took place in Luxembourg (in 2022, 2024, 2025) and one in Hangzhou (in 2023).

Invited speakers:

• Juan Duran (TU Delft)
• Alex Grzankowski (KCL)
• Nina Poth (Radboud University)
• Matthieu Queloz (Bern)
• Kate Vredenburgh (LSE)

Up to 4 contributing speakers will be selected through an open call.