skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


This content will become publicly available on May 1, 2026

Title: The transparency dilemma: How AI disclosure erodes trust
As generative artificial intelligence (AI) has found its way into various work tasks, questions about whether its usage should be disclosed and the consequences of such disclosure have taken center stage in public and academic discourse on digital transparency. This article addresses this debate by asking: Does disclosing the usage of AI compromise trust in the user? We examine the impact of AI disclosure on trust across diverse tasks—from communications via analytics to artistry—and across individual actors such as supervisors, subordinates, professors, analysts, and creatives, as well as across organizational actors such as investment funds. Thirteen experiments consistently demonstrate that actors who disclose their AI usage are trusted less than those who do not. Drawing on micro-institutional theory, we argue that this reduction in trust can be explained by reduced perceptions of legitimacy, as shown across various experimental designs (Studies 6–8). Moreover, we demonstrate that this negative effect holds across different disclosure framings, above and beyond algorithm aversion, regardless of whether AI involvement is known, and regardless of whether disclosure is voluntary or mandatory, though it is comparatively weaker than the effect of third-party exposure (Studies 9–13). A within-paper meta-analysis suggests this trust penalty is attenuated but not eliminated among evaluators with favorable technology attitudes and perceptions of high AI accuracy. This article contributes to research on trust, AI, transparency, and legitimacy by showing that AI disclosure can harm social perceptions, emphasizing that transparency is not straightforwardly beneficial, and highlighting legitimacy’s central role in trust formation.  more » « less
Award ID(s):
1943688
PAR ID:
10597787
Author(s) / Creator(s):
;
Publisher / Repository:
ScienceDirect
Date Published:
Journal Name:
Organizational Behavior and Human Decision Processes
Volume:
188
Issue:
C
ISSN:
0749-5978
Page Range / eLocation ID:
104405
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; (, Proceedings of Machine Learning Research)
    Matni, N.; Morari, M.; Pappas, G. J. (Ed.)
    We address the problem of learning the legitimacy of other agents in a multiagent network when an unknown subset is comprised of malicious actors. We specifically derive results for the case of directed graphs and where stochastic side information, or observations of trust, is available. We refer to this as “learning trust” since agents must identify which neighbors in the network are reliable, and we derive a learning protocol to achieve this. We also provide analytical results showing that under this protocol i) agents can learn the legitimacy of all other agents almost surely, and ii) the opinions of the agents converge in mean to the true legitimacy of all other agents in the network. Lastly, we provide numerical studies showing that our convergence results hold for various network topologies and variations in the number of malicious agents. 
    more » « less
  2. ; ; ; (, Evolution and Human Behavior)
    The concept of TRUSTWORTHINESS plays a role in the formation, maintenance, and dissolution of friendships, marriages, and cooperative relationships from small to large scales. Here, we analyze TRUSTWORTHINESS under the assumption that such concepts evolved to guide action adaptively. Intuition and research suggest that actors trust targets who have not engaged in betrayals. However, this perspective fails to capture certain real-world behaviors (e.g., when two people cheating on their spouses enter a relationship with each other and expect mutual fidelity). Evolutionary task analysis suggests that TRUSTWORTHINESS is structured to help actors address challenges of extending trust, where actors may gain or lose from doing so. In six experiments with American adults (N=1,718), we test the hypothesis that TRUSTWORTHINESS tracks not only (i) whether targets refrain from betraying trust when given opportunities, but also (ii) the impact of betrayal on the actor. Data generally support this hypothesis across relationships (friendships, romantic, professional): Actors deem non-betrayers more trustworthy than betrayers, but also deem betrayers more trustworthy when betrayals benefit actors. TRUSTWORTHINESS may incline actors to trust to those who refrain from betraying others—a potent signal of reluctance to betray oneself—while also favoring those who betray others if it serves oneself. 
    more » « less
  3. ; ; ; (, 5th Annual Conference on Learning for Dynamics and Control)
    We address the problem of learning the legitimacy of other agents in a multiagent network when an unknown subset is comprised of malicious actors. We specifically derive results for the case of directed graphs and where stochastic side information, or observations of trust, is available. We refer to this as “learning trust” since agents must identify which neighbors in the network are reliable, and we derive a protocol to achieve this. We also provide analytical results showing that under this protocol i) agents can learn the legitimacy of all other agents almost surely, and that ii) the opinions of the agents converge in mean to the true legitimacy of all other agents in the network. Lastly, we provide numerical studies showing that our convergence results hold in practice for various network topologies and variations in the number of malicious agents in the network. Keywords: Multiagent systems, adversarial learning, directed graphs, networked systems 
    more » « less
  4. ; (, American Sociological Review)
    When people want to conduct a transaction, but doing so would be morally disreputable, they can obfuscate the fact that they are engaging in an exchange while still arranging for a set of transfers that are effectively equivalent to an exchange. Obfuscation through structures such as gift-giving and brokerage is pervasive across a wide range of disreputable exchanges, such as bribery and sex work. In this article, we develop a theoretical account that sheds light on when actors are more versus less likely to obfuscate. Specifically, we report a series of experiments addressing the effect of trust on the decision to engage in obfuscated disreputable exchange. We find that actors obfuscate more often with exchange partners high in loyalty-based trustworthiness, with expected reciprocity and moral discomfort mediating this effect. However, the effect is highly contingent on the type of trust; trust facilitates obfuscation when it is loyalty-based, but this effect flips when trust is ethics-based. Our findings not only offer insights into the important role of relational context in shaping moral understandings and choices about disreputable exchange, but they also contribute to scholarship on trust by demonstrating that distinct forms of trust can have diametrically opposed effects. 
    more » « less
  5. ; ; (, Journal of Computing in Civil Engineering)
    This study aimed to investigate the key technical and psychological factors that impact the architecture, engineering, and construction (AEC) professionals’ trust in collaborative robots (cobots) powered by artificial intelligence (AI). This study seeks to address the critical knowledge gaps surrounding the establishment and reinforcement of trust among AEC professionals in their collaboration with AI-powered cobots. In the context of the construction industry, where the complexities of tasks often necessitate human–robot teamwork, understanding the technical and psychological factors influencing trust is paramount. Such trust dynamics play a pivotal role in determining the effectiveness of human–robot collaboration on construction sites. This research employed a nationwide survey of 600 AEC industry practitioners to shed light on these influential factors, providing valuable insights to calibrate trust levels and facilitate the seamless integration of AI-powered cobots into the AEC industry. Additionally, it aimed to gather insights into opportunities for promoting the adoption, cultivation, and training of a skilled workforce to effectively leverage this technology. A structural equation modeling (SEM) analysis revealed that safety and reliability are significant factors for the adoption of AI-powered cobots in construction. Fear of being replaced resulting from the use of cobots can have a substantial effect on the mental health of the affected workers. A lower error rate in jobs involving cobots, safety measurements, and security of data collected by cobots from jobsites significantly impact reliability, and the transparency of cobots’ inner workings can benefit accuracy, robustness, security, privacy, and communication and result in higher levels of automation, all of which demonstrated as contributors to trust. The study’s findings provide critical insights into the perceptions and experiences of AEC professionals toward adoption of cobots in construction and help project teams determine the adoption approach that aligns with the company’s goals workers’ welfare. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email