As algorithmic systems increasingly provide expert advice in various domains, questions arise about how audiences perceive such advice in terms of trustworthiness, credibility, and empathy. This study investigates how these perceptions are influenced by the source label (male expert, female expert, or AI system) and the communication style (neutral vs. empathic). Using a between-subjects experimental design, participants read a short advisory text concerning either sustainable pet ownership or sustainable investing. The content of the advice remained identical across all conditions; only the tone and attributed source label were systematically varied. After reading the text, participants evaluated the advisor on perceived trustworthiness, empathy, and credibility and indicated their intention to adopt more sustainable behaviours.

Preliminary analyses examine how the combination of source identity and communication style shapes audience perceptions of expertise and reliability. The study further explores whether empathic language can mitigate potential biases toward certain sources, particularly AI-generated advice, and whether effects differ across topic domains or author gender labels.

The study’s findings shed light on how subtle linguistic and contextual factors influence the reception of expert advice and how that effects their willingness to adopt sustainable practices. Insights from this research can inform the design of responsible and trustworthy communication strategies in both human and AI-mediated contexts, particularly for promoting engagement and behavioural change in sustainability-related decisions. More broadly, the study contributes to ongoing discussions about trust in expert systems and the ethical implications of presenting AI as a credible and empathetic advisor.