Digital human anchors are virtual entities generated through computer graphics (CG), artificial intelligence (AI), and associated technologies designed to exhibit anthropomorphic characteristics and interactive capabilities (Cui and Liu, 2023). As “mediators” between information sources and audiences, these anchors operate within various digital platforms, including online media. Their behavior can be driven either by real-time human operators or governed by algorithms and predefined scripts (Chakareski et al., 2022; Miao et al., 2022). Typically rendered as 2D or 3D models, they convey information through a combination of visual representations, speech, and nonverbal cues, showcasing progressively more sophisticated capabilities in semantic understanding and emotion simulation (Chakareski et al., 2022; Miao et al., 2022).

The current research on digital human anchors primarily addresses two key dimensions: anthropomorphic representation and social role enactment. Mindful of the “Uncanny Valley effect” and its potential influence on audience acceptance, researchers have proposed specific guidelines for the optimal visual design of digital human anchors. Studies informed by cognitive conflict theory suggest that attributes such as natural skin tones, realistic body proportions, and distinct gender markers are crucial for fostering positive perceptions of digital humans and their portrayed roles (Schwind et al., 2018). Furthermore, empirical investigations have demonstrated that artificial lip or hand movements can significantly increase negative audience reactions. This response aligns with perceptual mismatch theory and underscores the need to refine the naturalness of facial expressions, lip synchronization, and other “high sensitivity deficits” to improve the acceptability of digital human anchors (Wu et al., 2024). Regarding application domains, digital human anchors are commonly categorized into e-commerce, entertainment, and media sectors (Huang and Yu, 2023; Xu, 2023). For instance, in e-commerce, research indicates that digital human anchors may be more effective than their human counterparts in certain contexts, such as promoting specific products (Yan et al., 2025) or conveying discount information (Chang et al., 2025), to stimulate consumer willingness. Within the entertainment sphere, virtual applications such as Replika exemplify how entertainment anchors, leveraging intelligent voice and real-time interaction technologies, can offer users highly immersive conversational experiences that simulate interactions with a real person (Hu et al., 2021; Laestadius et al., 2024).

In summary, the existing research on digital human anchors has yielded initial findings, predominantly within e-commerce and entertainment, highlighting their potential applications and user preferences in these commercial settings. Nevertheless, a significant portion of the literature prioritizes commercial value, leaving the role and impact of digital human anchors in public affairs communication, particularly public policy communication, underexplored. Given that public-policy communication requires high levels of trust and understanding, the comparative efficacy of digital versus human anchors in this domain warrants immediate investigation. To address this gap, the present study, situated within the public-policy communication context, aims to compare the effects of communicator type (digital human vs. human anchor) on policy adoption intentions. Furthermore, it investigates the potential mediating role of psychological distance and the moderating effects of varying credibility endorsements.

Existing studies have suggested that digital avatars may be less effective than human hosts in terms of communication and persuasion. Communication theory identifies social presence as a critical factor shaping communication outcomes. Media with a higher social presence (e.g., live videos) tend to generate stronger emotional connections and behavioral responses from audiences than those with a low presence (Wang et al., 2024). In tandem, perceived authenticity is essential. Research by Nicolette showed that the inherent authenticity of human hosts helps prevent audience rejection, leading to greater acceptance and recognition that virtual avatars cannot fully replicate (Filova et al., 2025). Additionally, source credibility is a key driver of persuasion (Li et al., 2023). Li found that human hosts, viewed as more authoritative and credible, are more effective than virtual avatars in influencing brand purchase intentions. Since public-policy communication similarly depends on source credibility to facilitate information acceptance and behavior change, this study extends this finding to policy adoption by hypothesizing that human presenters have similar advantages in these contexts.

Psychological distance denotes the perceived separation between an individual’s self and an object (e.g., event, person, message, encompassing). It encompasses dimensions such as temporal, spatial, social, and hypothetical distances. Construal Level Theory (CLT) posits that psychological distance influences not only cognitive construal, but also significantly impacts attitudes and behavioral intentions. Research has demonstrated that psychological distance can modulate trust (Bandara et al., 2021), affect consumer choice (Goodman and Malkoc, 2012), and shape purchase intentions through its effect on information perception (Liu et al., 2021). Within public-policy communication, a reduced psychological distance toward the information source or message is expected to lead to more favorable information processing; heightened perceived relevance; and increased understanding, agreement, and adoption willingness.

Communicator type is a key determinant of perceived psychological distance. Human anchors generally elicit a higher social presence and greater perceived authenticity (Park, 2024) than digital anchors. Consequently, audiences may perceive human anchors as socially nearer and hypothetically more tangible, reducing the overall psychological distance. This study anticipates that the difference in psychological distance, contingent on communicator type, will subsequently influence the willingness to adopt public policy. Thus, psychological distance is proposed to mediate the effect of communicator type on policy adoption.

With the growing pervasiveness of human-computer interaction (HCI), anthropomorphism, which is defined as the attribution of human-like characteristics, intentions, or behaviors to nonhuman entities, has emerged as a central concept in HCI research.

Extensive research has indicated that enhancing the anthropomorphism of digital entities generally yields positive outcomes. For instance, in marketing, anthropomorphic chatbots, advertisements, and brand mascots have proven effective in increasing consumer purchase intentions, favorable advertising attitudes, and psychological intimacy (Han, 2021; Ko et al., 2022). In HCI and user acceptance studies, highly anthropomorphic digital humans or bots tend to facilitate more efficient interactions (Lv et al., 2023), garner greater user trust and affective acceptance, and promote better understanding, particularly in service recovery contexts (Frazer, 2022).

These positive effects may be partly attributable to the capacity of anthropomorphism to reduce psychological distance. When digital entities exhibit a high degree of similarity to humans in appearance, vocal characteristics, and demeanor (Frazer, 2022), audiences may perceive them as socially “closer” and hypothetically more “real.” These attributes, in turn, can lessen the perceived disconnect that is often experienced in human-machine interactions.

Consequently, we theorize that the degree of anthropomorphism exhibited by digital anchors may modulate the effect of communicator type (human versus digital) on the public’s perceived psychological distance. Based on this reasoning, the following hypothesis is proposed:

Credibility endorsement is the process of bolstering the perceived credibility of a message or information source through external cues such as endorsements from authoritative entities or evidence of broad social acceptance. This process serves as a key mechanism for effective dissemination of information.

As a significant cognitive heuristic (Mena et al., 2020), credibility endorsement enables audiences to rapidly evaluate source or message credibility, particularly in situations of information overload or uncertainty, by relying on cues such as reputation, consistency, and social approval (Metzger and Flanagin, 2013). The efficacy of this mechanism has been established across diverse domains. However, its impact can vary depending on factors such as product type (e.g., experiential versus search goods) and the nature of the endorser (e.g., platform-based endorsements such as Netflix recommendations versus traditional celebrity endorsements).

Within the framework of this study, we propose that external credibility endorsements exert a second-order moderating influence (or re-moderating effect) on the previously discussed moderating effect of anthropomorphism (i.e., the degree to which anthropomorphism alters the impact of anchor type on psychological distance). The underlying rationale is that the strength of external credibility cues can shift the audience’s reliance away from the communicator’s intrinsic characteristics (such as their human versus non-human nature and degree of anthropomorphism) toward these external signals, thereby influencing the perceived psychological distance.

Therefore, the following hypothesis is proposed.

To enhance external validity, students were excluded from the formal experiments. This exclusion stems not solely from the “plastic pollution control” topic, but from widespread findings in social science and behavioral research that student groups exhibit high structural homogeneity in age, educational background, and other factors (Gallander Wintre et al., 2001), Their social roles remain fluid, their connection to public policy interests is relatively weak, and their attitudes are more susceptible to fluctuation due to conformity and peer influence (Peterson and Merunka, 2014). These characteristics may weaken the generalizability of the conclusions drawn in this study. Therefore, student samples were excluded from the formal experiment for methodological reasons.

This study employed the professional statistical software G*Power for the preliminary sample size calculation and analysis to ensure the accuracy of the experimental effects and stability of the statistics. Considering a moderate effect size of 0.25, significance level of 0.05, and power of 0.8, studies 1, 2, and 3 were projected to require sample sizes of 128, 158, and 270, respectively. Study 1 collected 128 questionnaires; Study 2 collected 241 questionnaires; and Study 3 collected 470 questionnaires, totaling 839.

Regarding demographic characteristics, the sample exhibited a balanced distribution across sex, age, occupation, and educational background, reflecting the general public profile. This distribution supported the reliability and interpretability of the experimental results (Table 1).

This study selected “plastic pollution control” as its experimental topic based on three primary considerations: First, plastic pollution represents a quintessential environmental governance issue, involving ecological sustainability, social responsibility, and economic equilibrium. It constitutes a complex policy context characterized by an interdisciplinary nature and the interplay of multiple stakeholders. Compared to topics such as taxation and personal health, plastic pollution control combines public relevance with urgency, making the topic more likely to spark widespread audience attention and discussion. These factors facilitate a better understanding of, and engagement with, the issue across diverse cultural backgrounds. Second, environmental policies are often regarded as “marginal.” During policy adoption, they minimize interference from existing stances and political ideologies, allowing audiences to focus more intently on the policy’s substance and communication methods (Guston, 2001). Therefore, using environmental governance issues as the experimental material ensures that participants from diverse cultural backgrounds engage in discussions within a similar framework, enhancing the contextual relevance of the experiment and the interpretive power of its results. Finally, tackling plastic pollution has become a key priority in China’s governance efforts, exemplified by the issuance of the “Action Plan for Plastic Pollution Control during the 14th Five-Year Plan Period.” This official focus elevates the policy issue to a high practical relevance, provides a concrete policy context for cultural adaptation, and strengthens the real-world significance of this study.

To ensure content validity, all scales in this study were designed by referencing established and mature instruments. The research comprises four variables and 14 items using a 7-point Likert scale, as shown in Table 2.

This study systematically investigates the efficacy of emerging digital anchors in public policy communication by examining the impact of the communicator type on public policy adoption intentions. It further analyzes the complex interplay among psychological distance, anthropomorphism, and credibility endorsement. Based on the empirical findings, the primary conclusions are as follows.

First, human anchors demonstrate a significant advantage in fostering policy adoption, with psychological distance emerging as a critical mediating mechanism. Second, the effectiveness of digital anchors is contingent on context, with the degree of anthropomorphism acting as a significant moderating factor. Third, the influence of external cues is substantial, with credibility endorsement demonstrating a second-order moderating effect (i.e., moderated moderation) on the impact of anthropomorphism. The details are listed in Table 6.

Contextualized by the increasing integration of digital technology within public governance, this study systematically investigates the differential communicative efficacy of human versus digital anchors in policy dissemination and their underlying mechanisms of action. By employing a rigorous online experimental design, this study offers novel cognitive insights and empirical evidence for pertinent theoretical fields. The specific theoretical contributions of this study are as follows:

First, it significantly broadens the existing academic agenda by extending the investigation of digital communicators to the critical domain of public policy communication. Prior scholarship has predominantly explored digital human applications in commercial contexts, such as marketing (Han et al., 2023) and customer service or service recovery (Liu et al., 2023). Given the scarcity of current research focusing on the field of public policy communication, our findings not only corroborate the current general advantage of human anchors in fostering policy adoption intentions, but also reveal the potential for high-anthropomorphism digital anchors to mitigate this efficacy gap under specific circumstances, such as when paired with strong credibility endorsements. This contribution addresses a theoretical lacuna at the nexus of digital communication technology, public administration, and political communication while also providing crucial insights for delineating the appropriate application boundaries for digital humans in public affairs.

Second, this research enhances the comprehension of social cognitive mechanisms and the role of psychological distance in HCI. Moving beyond prior studies that often centered on specific emotional responses such as empathy (Liu-Thompkins et al., 2022), humor (Xu and Liu, 2022), or apology (Song et al., 2023), the current work concentrates on the pivotal dimension of perceived psychological distance toward different communicators. It confirms the critical mediating function of psychological distance in the relationship between communicator type and policy adoption intentions and innovatively reveals how the level of anthropomorphism moderates this psychological process. These findings indicate that high levels of anthropomorphism effectively reduce perceived psychological distance by closely simulating human characteristics. They advance anthropomorphism theory by emphasizing static feature descriptions to understand dynamic social perceptions in interactive contexts. Moreover, these results underscore anthropomorphism’s influence on superficial user experience and profound impacts on audience cognition, attitudes, and behavioral intentions, through the mechanism of reshaping psychological distance.

Furthermore, this study highlights the distinct value of credibility endorsement in modulating the efficacy of novel communicators in the digital era, and promotes the integration of relevant theories. The empirical results demonstrate that strong credibility endorsements significantly augment the communication effectiveness of digital human anchors, particularly those with high anthropomorphism. The underlying mechanism suggests that potent external credibility signals can mitigate or neutralize the inherent limitations of digital communicators concerning perceived authenticity and social proximity, thereby influencing the manifestation of anthropomorphic effects. This finding enriches the theoretical discourse on reputation governance in public management (Pedersen and Salomonsen, 2023), illustrates credibility endorsement as a crucial link among organizations, interactive technologies, and the public, and offers new perspectives for public sector credibility management amidst digital transformation. Concurrently, this study integrates and empirically validates the key tenets of persuasion theory within digital policy communication: communicator type (source attributes), anthropomorphism (message presentation style), and credibility endorsement (contextual cues) collectively influence audience attitudes and intentions. This integrated multifactorial approach advances the development of a comprehensive theoretical framework for understanding complex policy communication dynamics in the digital era, moving beyond single-factor analyses.

This study’s findings not only enrich the relevant theoretical knowledge, but also offer targeted practical insights for the public sector to effectively leverage digital anchors for policy communication in an increasingly digitized environment.

Research has confirmed that human anchors retain a broad advantage in fostering policy adoption, especially when deep trust and emotional connections need to be established. For major complex policies or those involving significant adjustments to public interests, senior human anchors should be prioritized for authoritative interpretation to leverage their unique values in professional expertise, empathy, and trust building. For routine, high-frequency tasks, such as information dissemination, standard Q&A, and service guidance, digital anchors (especially highly anthropomorphic types) can be deployed to enhance efficiency and coverage. In China, cities such as Shenzhen and Chengdu have achieved notable success in disseminating carbon neutrality and energy conservation policies through the introduction of digital hosts. More importantly, regions can advance human-machine collaboration in tiers based on their economic and technological capabilities. For instance, resource-rich central cities can adopt a “human-led, digital-assisted” model for major policy announcements and complex issue communications, with human anchors taking the lead in in-depth explanations. Concurrently, digital anchors provide multimodal background materials, terminology explanations, or real-time interactive Q&A to enhance credibility. This complementary approach transforms communication from “broad coverage” to “targeted precision,” to achieve synergistic dissemination.

This study highlights the critical role of anthropomorphism in regulating the effectiveness of digital anchors, with highly anthropomorphic ones significantly narrowing the psychological distance between them and their audiences. Therefore, the practical application of digital human anchors should not merely aim for superficial resemblance, but also pursue a deeper sense of authenticity. Resources must be invested in continually enhancing the multimodal expressiveness of digital anchors, including more natural speech intonation, richer facial expressions, and more coordinated body movements. Additionally, exploring the integration of affective computing technology is essential for endowing them with situational awareness and emotional interaction capabilities. The goal is to make digital hosts perceive themselves as “socially acting entities” rather than as cold machines, thereby effectively reducing psychological barriers among the public and enhancing the smoothness and receptiveness of information delivery.

A key finding of this study is the significant second-order moderating effect of strong credibility endorsement, which effectively narrows the perceived psychological distance gap between high-anthropomorphism digital and human anchors. This effect underscores the critical importance of prioritizing and effectively leveraging credibility endorsements when promoting the use of digital avatars for policy communication. Public sector entities must provide clear and authoritative identity verification and source attributes for digital avatars (Li and Yang, 2024). For instance, they can be integrated into official media platforms or government portals through endorsements or certifications from the relevant authoritative bodies. Such robust external signals can significantly enhance the credibility of digital hosts, compensating for their inherent limitations in authenticity and social presence, thereby increasing public trust in and acceptance of the information they convey. However, strong credibility endorsements are merely a “steppingstone” for establishing initial public trust. Only through professional cultivation of knowledge dissemination, emotional interaction, and value expression can digital hosts better showcase the unique personality and value of this new communication entity. These factors enable the public to “trust because of the content” rather than “trust because of authority,” achieving the leap from being trusted to being trustworthy.

Optimizing “anthropomorphic design” and effectively leveraging “authoritative credibility endorsements” as key strategies to enhance the communication effectiveness of digital avatars carry potential ethical risks and governance challenges. Therefore, establishing robust governance standards are necessary and urgent. To ensure the responsible application of digital avatars in policy communication, comprehensive technical standards (Batool et al., 2025), application scenarios (Park, 2024), and guiding principles or norms for ethical constraints (García de Torres et al., 2025) should be promptly formulated. Among these, an ethical responsibility exists to safeguard the public’s right to know through mandatory, clear “AI-generated” identity markers (Veale and Zuiderveen Borgesius, 2021). Avoiding confusion among real individuals is fundamental to maintaining transparency and securing the public’s right to know. Simultaneously, clearly defining appropriate application scenarios (such as policy interpretation and public services), strictly prohibiting overreach, and establishing content reviews and public feedback mechanisms are essential for mitigating risks, facilitating continual improvement, and fostering public trust (Yuan and Chen, 2025).