AI Anchors as Transmitters of Public Diplomacy: Perceived Credibility of Environmental and Political News in China

In recent years, major news organizations worldwide have increasingly adopted artificial intelligence (AI) powered news anchors as potential tools for the delivery of news content, for example, India has Sana and Lisa, Greece has Hermes, Kuwait has Fedha and Taiwan has Ni Zhen.

With the ongoing evolution of this trend, scholars and journalistic professionals are beginning to investigate the capacity of AI to address current difficulties in this field, particularly concerning perceived credibility. This includes state-run news agencies that utilize political news to promote public diplomacy, while environmental news is also often included due to the strong connection between climate change and public diplomacy (Song, 2025). Given that some view AI as less biased and more accurate than human journalists (Sundar, 2008), there are suggestions that AI could help improve perceptions of bias in public diplomacy efforts.

A notable example of the use of AI anchors is Xinhua News Agency, a state-run news agency in China that serves as a key channel for information dissemination by the Chinese government. The agency introduced AI anchors in 2018 and the world’s first female AI news anchor in 2019. With the continuous advancements in AI technology, the newest generation of AI news anchors is now being used on various platforms, including social media sites such as YouTube and X.

While the potential of AI anchor is widely recognized, it raises concerns about the prevalence of state-sanctioned narratives (Bohacek & Farid, 2024) and issues related to credibility. Xinhua's AI anchors exhibit characteristics echoing such concerns: their physical appearances are modeled after real journalists from the agency, it does not always introduce clear labels indicating that they are powered by AI, making it very difficult to distinguish between human presenters and AI presenters.

Therefore, this study is planned to examine how the Chinese government uses AI anchors to promote public diplomacy and focus on audience acceptance of AI anchors within international communication to explore whether the incorporation of AI anchors can improve the effectiveness of public diplomacy. This study aims to make three theoretical contributions through two key variables: AI label and machine heuristic. First, the analysis will explore how the machine heuristic mediates the relationship between political and environmental news and perceived credibility. Second, both machine heuristic and perceived credibility will be moderated by the presence of an AI label (Label vs. No label). Finally, the study will demonstrate that the perceived credibility of public diplomacy content conveyed by AI anchors, mediated by the machine heuristic, is further influenced by the presence or absence of a label.

Public Diplomacy and News Type
Public diplomacy can be linked to the concept of “soft power”, popularised by Joseph Nye in the 1980s, which is often considered to be promoted through public diplomacy rather than propaganda as the latter is often associated with issues of credibility. In China, “soft power” has become a widely used term, However, Min and Luqiu (2021) argue that, at its core, this is simply a globalization of the use of propaganda under the guise of soft power. Reflecting the fact that the transition from international propaganda to public diplomacy has not yet been fully realized. China's soft power campaigns have therefore had limited success in persuading international audiences. To adapt to the changing media environment, China's public diplomacy framework has undergone a series of changes. Previous studies (Min & Luqiu, 2021) have found that a soft news-centered production model is emerging, eschewing the didactic, rigid, and hard-line approach to public diplomacy. State-run news media now place greater emphasis on soft news and commercialized forms of infotainment (Liang, 2019), frequently covering topics such as climate change. However, there is limited research on the impact of soft topics like climate change on public diplomacy, particularly when AI technologies are integrated into news delivery.

Machine Heuristic and Perceived Credibility
Sunder (2008) proposed the Modality-Agency-Interactivity-Navigability (MAIN) model to support this notion, arguing that technological affordances manifested as interface cues trigger specific cognitive heuristics that guide users in assessing the credibility of an information system and its content. One key heuristic identified in the MAIN model is the “machine heuristic”. According to this heuristic, when individuals perceive the source of interaction as a machine rather than a human, they tend to apply stereotypes about machines—viewing them as mechanical, objective, and ideologically unbiased.

This study uses the concept of machine heuristics as a mediator of the relationship between public diplomacy news provided by AI anchors and perceived trustworthiness. As Yang and Sunder (2024) point out, if the interface lacks a machine-like appearance, this may undermine the machine heuristic and lead to a reduction in credibility.

So the following research question has been proposed: (1) How does the machine heuristic mediate the relationship between political and environmental news delivered by AI anchors and their perceived credibility? (2) How do the indirect effects of news type (political vs. environmental news) on perceived credibility mediated through machine heuristic vary according to AI label?

Method
An online experiment will be conducted in the US using a 2x2 between-subjects design, examining the type of news (political news vs. environmental news) and labeling (labeled vs. non-labeled). Participants will view two short news segments from Xinhua News Agency presented by AI anchors: one covering a climate or environmental issue, and the other related to an international political news story. After viewing, participants will complete a questionnaire to assess their perceptions and responses.