Conference Agenda

Communicative AI is reshaping news consumption, yet audience perceptions remain understudied—particularly across the distinct roles AI plays as news creator and news converser. This systematic literature review (SLR), following PRISMA guidelines, synthesizes empirical, audience-centered research on how news users perceive, evaluate, and interact with communicative AI in both capacities. Drawing on Sundar and Lee's (2022) framework, we distinguish between AI as news creator and presenter (e.g., AI-generated articles, AI news anchors) and AI as news converser (e.g., interactive chatbots, conversational agents). Two separate search protocols were developed, yielding 68 studies for the first strand and 25 for the second. While audiences remain broadly skeptical of AI-generated news, younger users are increasingly integrating AI interfaces into everyday news practices. Key findings — including dominant research themes, methodological trends, and gaps in the literature — will be presented at the conference.

Political actors increasingly use social media to spread subtle climate disinformation questioning policy effectiveness rather than climate scienc—so-called response skepticism. This preregistered experiment (N = 1,304) compared factual corrections, two forms of source discreditation (track-record vs. intent-based), and their combinations as user-generated counter-interventions on social media. Only intent-based source discreditation reliably reduced misperceptions and response skepticism, outperforming factual correction on the latter. Combined interventions showed no added benefit. Mediation analyses indicated that intent-based discreditation worked by increasing perceived manipulative intent (PMI), which in turn reduced misperceptions and skepticism—consistent with the Persuasion Knowledge Model. These effects were not moderated by climate skepticism or party identification, suggesting PMI is a relatively bias-resistant mechanism. The findings highlight deceptive intent framing as a promising, scalable countermeasure against elite-driven climate disinformation, more effective than content-focused corrections in politically charged digital environments.

This study uses mixed-methods social network analysis to explore the structure of right-wing populist and conspiracy networks on public German Telegram groups. Analyzing data scraped between September 2024 and September 2025 via exponential snowball sampling, the study constructed a network of 15,959 nodes and 59,346 edges. Community detection identified 11 distinct clusters (Modularity Q=0.457). Qualitative coding of the most central channels revealed a pervasive right-wing populist presence and a significant thematic focus on pro-Russian war propaganda. Additionally, hyperlink analysis of 88,981 messages demonstrated how the network connects to broader online discourses. Findings indicate that distinct clusters fulfill specific functions—from soft entry points for conspiracies to protest mobilization—structurally facilitating a "false balance" of facts by completely omitting opposing mainstream viewpoints.

The popularity of ADHD-related social media content has raised concerns about self-diagnosis. Based on the SESAM model, which suggests that selective exposure to health-related media increases the salience of related health goals, this study examines the relationship between exposure to adult ADHD content on social media and self-diagnosis of the condition. An experiment using Instagram-like stimuli, some mentioning ADHD and some not, tested whether exposure to ADHD-related content influenced self-diagnosis among participants who had and had not previously self-diagnosed with ADHD. The mediating role of social comparison was also examined. Results showed that exposure to ADHD-related content increased post-exposure self-diagnosis scores among participants who had already self-diagnosed before exposure. No effect was found among participants without prior self-diagnosis. Social comparison did not mediate the relationship. However, it was a strong predictor of self-diagnosis.