Full Disclosure, Less Trust? How the Level of Detail about AI Use in News Writing Affects Readers' Trust

TL;DR

This study addresses how the level of detail in AI disclosures within news content shapes readers' trust and behavioral responses. Using a $3\times 2\times 2$ mixed factorial design, it manipulates disclosure level (none, one-line, detailed), news type (politics vs lifestyle), and AI involvement (low vs high) to measure trust, source-checking, and subscription decisions, supplemented by qualitative interviews. The findings show that detailed disclosures can reduce article-level trust and subscriptions, while one-line disclosures perform similarly to no disclosure; source-checking increases under both one-line and detailed disclosures. Interviews reveal that curiosity drives source-checking whereas trust determines subscription, suggesting a trade-off rather than an inexorable transparency dilemma. Practically, concise disclosures or detail-on-demand designs may balance readers’ transparency expectations with retention of trust, with policy implications for regulation and industry practice.

Abstract

As artificial intelligence (AI) is increasingly integrated into news production, calls for transparency about the use of AI have gained considerable traction. Recent studies suggest that AI disclosures can lead to a ``transparency dilemma'', where disclosure reduces readers' trust. However, little is known about how the \textit{level of detail} in AI disclosures influences trust and contributes to this dilemma within the news context. In this 3$\times$2$\times$2 mixed factorial study with 40 participants, we investigate how three levels of AI disclosures (none, one-line, detailed) across two types of news (politics and lifestyle) and two levels of AI involvement (low and high) affect news readers' trust. We measured trust using the News Media Trust questionnaire, along with two decision behaviors: source-checking and subscription decisions. Questionnaire responses and subscription rates showed a decline in trust only for detailed AI disclosures, whereas source-checking behavior increased for both one-line and detailed disclosures, with the effect being more pronounced for detailed disclosures. Insights from semi-structured interviews suggest that source-checking behavior was primarily driven by interest in the topic, followed by trust, whereas trust was the main factor influencing subscription decisions. Around two-thirds of participants expressed a preference for detailed disclosures, while most participants who preferred one-line indicated a need for detail-on-demand disclosure formats. Our findings show that not all AI disclosures lead to a transparency dilemma, but instead reflect a trade-off between readers' desire for more transparency and their trust in AI-assisted news content.

Figures (4)

• Figure 1: An example of one-line disclosure for a Low AI news article
• Figure 2: A screenshot of the token-spending decision task and the corresponding revealed source
• Figure 3: Overview of the experimental procedure. All participants had the no disclosure condition as the first session, while the order of low- and high-AI disclosure sessions (Sessions 2 and 3) was alternated and counterbalanced across participants. Each session consisted of four articles and associated questionnaires and ended with questions about the outlet that published them.
• Figure 4: Stacked bar graph visualization of (a) average media trust scores (article level) across the three disclosure conditions, and (b) average token spending behavior across the three disclosure conditions.