Addressing trust and distrust in artificial intelligence in newspaper texts

Abstract

Abstract Trust in new technologies is often seen as a key condition for their implementation in everyday life, as the aim is for users to utilise technical innovations without strong reservations. The aspect of trust also plays a fundamental role in the discourse on Artificial Intelligence (AI) and its far-reaching consequences for people’s lives. In this context, journalistic reporting as part of the public discourse plays a considerable role in creating trust or distrust in AI as a new technology. Whether and how trust issues are addressed in journalistic reporting has not been investigated so far. The paper explores how AI is depicted in German journalistic coverage, covering the topics of autonomous driving and smart home systems, which we expected to generate diverging trust issues. Furthermore, we show how trustworthiness and untrustworthiness are constructed in those texts, choosing AI applications intended for everyday use and therefore relying on trust from users ( Tschopp et al. 2022 ; Gillespie et al. 2021 ). We carried out a qualitative linguistic content analysis on a corpus of German press texts from 2023 that are dedicated to the aforementioned topics. We focused our analysis on explicit and implicit trust communication in the journalistic coverage of AI and examined whether qualitative differences can be identified in the way trust or distrust is addressed in various sub-discourses.