从“像人”到“可信”——AI虚拟主播的拟真性提升与新闻可信度构建的协同研究

From “Human-Likeness” to “Credibility”—A Synergistic Study on the Enhancement of Realism in AI Virtual Anchors and the Construction of News Credibility

摘要: 随着人工智能、动作捕捉以及语音合成等技术的飞速演进，传媒领域的生产逻辑与传播形态正在发生深刻变化。AI虚拟主播作为智能化的创新产物，已从早期的机械播报，演进为具备超写实外观、自然语音交互与动态表情的新媒介主体。国内媒体积极拥抱新技术，推出如新华社的“新小萌”、央视“王冠”等专属AI虚拟主播，并使其参与到重大事件报道中，应用场景覆盖时政新闻、民生播报、赛事解说等各领域。这一技术的应用不仅重构了新闻生产的“人机协同”模式，更引发关于AI虚拟主播的拟真性与新闻可信度之间的讨论。“像人”与“可信”之间的关系并非自然递进，而是一种充满辩证色彩的协同演进。拟真性的增强既为构建新闻可信度提供了机遇，也带来恐怖谷效应、伦理失范与真实性消解等严峻挑战。因此，本研究旨在系统梳理AI虚拟主播拟真性与新闻可信度之间的协同逻辑，剖析二者协同发展的现实困境，并提出针对性的优化路径，力图为新闻传播在人工智能时代的持续发展提供理论参考。

Abstract: With the rapid advancement of artificial intelligence, motion capture, and speech synthesis technologies, the production logic and communication structures of the media industry are undergoing profound transformations. As an innovative product of intelligent technologies, the AI virtual anchor has evolved from early mechanical broadcasting to a new type of media agent characterized by hyper-realistic appearance, natural speech interaction, and dynamic facial expressions. Chinese media organizations have actively embraced these emerging technologies by launching exclusive AI virtual anchors—such as Xinhua News Agency’s Xin Xiaomeng and China Central Television’s Wang Guan—and deploying them in the coverage of major events. Their applications now span political news, public affairs reporting, sports commentary, and other fields. The adoption of AI virtual anchors not only reconstructs the “human–machine collaboration” model of news production but also prompts renewed discussions on the relationship between virtual-anchor realism and news credibility. The connection between “human-likeness” and “trustworthiness” is not a natural or linear progression; rather, it represents a dialectical process of synergistic evolution. While enhanced realism offers opportunities for strengthening news credibility, it also introduces significant challenges, including the uncanny valley effect, ethical risks, and threats to authenticity. Therefore, this study aims to systematically examine the synergistic logic between AI anchor realism and news credibility, analyze the practical dilemmas in their co-development, and propose targeted optimization strategies, with the goal of providing theoretical guidance for the sustainable development of journalism in the era of artificial intelligence.