Generative Artificial Intelligence Empowering Propaganda, Ideology, and Cultural Work: Evolution, Risks, and Responses

“Propaganda, ideology, and cultural work concerns the future and destiny of the Party, concerns the long-term stability and security of the country, concerns national cohesion and centripetal force, and is an extremely important task.” [1]1 A reference to the concern that in human-machine systems, control may shift from humans to machines, a theme in discussions of technological autonomy. At present, the world is undergoing changes unseen in a century accelerating in their evolution, a new round of technological revolution and industrial transformation is deepening, and propaganda, ideology, and cultural work has entered a new stage of development. Generative artificial intelligence refers to technologies that generate content such as text, images, sound, video, and code based on algorithms, models, and rules. Unlike the analytical functions of traditional artificial intelligence, generative artificial intelligence can learn and generate new content with logic. In recent years, with the vigorous development of artificial intelligence technology, especially represented by ChatGPT and DeepSeek, major changes have occurred in the field of generative artificial intelligence, triggering a global new wave of generative artificial intelligence penetrating various fields, with the user scale showing extremely rapid growth. However, with the rapid iteration of artificial intelligence technology and the increasing complexity of the network environment, while generative artificial intelligence brings convenience, it also harbors many risks and negative effects. Therefore, deeply grasping the new characteristics of the development of generative artificial intelligence tools, clarifying their internal influence mechanisms, and enhancing the adaptability, competitiveness, and vitality of propaganda, ideology, and cultural work in the era of generative artificial intelligence have become an urgent task.

“宣传思想文化工作事关党的前途命运，事关国家长治久安，事关民族凝聚力和向心力，是一项极端重要的工作。”当前，世界百年未有之大变局加速演进，新一轮科技革命和产业变革深入发展，宣传思想文化工作进入新发展阶段。生成式人工智能是指基于算法、模型、规则生成文本、图片、声音、视频、代码等内容的技术，不同于传统人工智能的分析功能，生成式人工智能可以学习并生成具有逻辑的新内容。近年来，随着人工智能技术蓬勃发展，尤其是以ChatGPT和DeepSeek为代表，生成式人工智能领域发生重大变革，在全球引发了生成式人工智能渗透各领域的新浪潮，用户规模呈极速增长态势。然而，随着人工智能技术的快速迭代与网络环境的日益复杂，生成式人工智能在带来便捷的同时，也潜藏着诸多风险与负面效应。因此，深刻把握生成式人工智能工具发展的新特性，厘清其内在影响机理，增强生成式人工智能时代宣传思想文化工作的适应力、竞争力和生命力，已成为一项紧迫任务。

I. Behind the Hype: New Characteristics of Generative Artificial Intelligence Empowering Propaganda, Ideology, and Cultural Work

一、爆红背后：生成式人工智能赋能宣传思想文化工作的新特性

Generative artificial intelligence is a strategic technology and revolutionary force leading a new round of technological and industrial revolution. Writing imaginative poetry in multiple languages, quickly generating paper abstracts, and summarizing the week's major events into a news review demonstrate the ability of generative artificial intelligence to solve various general-purpose problems. Only by actively embracing the beneficial achievements of scientific and technological development can we promote the innovative development of propaganda work. The embedding of generative artificial intelligence into propaganda, ideology, and cultural work brings new developments to it.

生成式人工智能是引领新一轮科技产业革命的战略性技术和革命性力量。多语言撰写充满想象力的诗歌、快速生成论文摘要、将一周大事写成新闻综述等，显示出生成式人工智能解决多种通用问题的能力。积极拥抱科学技术发展的有益成果，才能推动宣传工作创新发展。生成式人工智能嵌入宣传思想文化工作为其带来新发展。

(I) The “Co-frequency” Pattern: Rich and Efficient Content Supply

（一）“共频”样态：内容供给丰富化与高效化

In the era of generative artificial intelligence, the content supply of propaganda, ideology, and cultural work is high-speed and diverse, with supply efficiency beyond imagination, presenting a pattern of “resonating with technology and co-frequency with the demand side” on the supply side of propaganda, ideology, and cultural work. Generative artificial intelligence can complete tasks including instant chat, writing emails or papers, coding, producing video scripts and copy, and translation, closely related to learning and work content. In the context of media convergence, artificial intelligence technology has penetrated key links such as propaganda content production. For example, the “free canvas” jointly launched by Baidu Wenku and Baidu Wangpan supports one-click sharing and storage of rich media documents (information documents presented in dynamic, multimedia forms on the internet), covering the entire process from content production to content consumption. Unlike traditional artificial intelligence products that focus more on a specific vertical professional field, the super-strong content production capability of generative artificial intelligence makes it suitable for the propaganda field. The generative artificial intelligence model cannot be separated from the support of big data in the era of the digital economy. Compared with traditional artificial intelligence, the diversified massive data mastered by generative artificial intelligence gives it a super-large knowledge base, making it an important way for people to obtain information. It also makes the content sources of propaganda, ideology, and cultural work more extensive, which will far exceed the knowledge scope of individual propaganda workers or groups, thereby becoming a living source of propaganda content. Coupled with continuous technological iteration and optimization and upgrading, large language models make generative artificial intelligence increasingly intelligent and efficient, further lowering the threshold for user use. For example, by trying to give a few keywords, generative artificial intelligence can easily write a news draft of several hundred words with characters, events, and viewpoints in just over ten seconds, greatly increasing the supply speed of propaganda works.

生成式人工智能时代，宣传思想文化工作的内容供给高速、多样，供给效率超乎想象，使宣传思想文化工作供给侧呈现出“与科技共鸣、与需求端共频”的样态。生成式人工智能可以完成包括即时聊天、撰写邮件或论文、编写代码、制作视频脚本文案和翻译等任务，与学习、工作内容紧密相关。在媒介融合的背景下，人工智能技术已渗入宣传内容生产等关键环节。例如，百度文库和百度网盘联合推出的“自由画布”支持对富媒体文档（互联网上以动态、多媒体形式呈现的信息文档）的一键分享和存储，覆盖从内容生产到内容消费的全流程。与传统人工智能产品更专注于某一垂直专业领域不同，生成式人工智能的超强内容生产能力使其适用于宣传领域。生成式人工智能模型的背后离不开数字经济时代大数据的支持。相较于传统人工智能，生成式人工智能掌握的多元化海量数据使其拥有超大型知识库，成为人们获取信息的重要方式。它也让宣传思想文化工作的内容来源更加广泛，这将远远超过宣传工作者个体或群体的知识涉猎面，继而成为宣传内容的源头活水。加之技术不断迭代与优化升级，大语言模型使得生成式人工智能变得愈发智能、高效，进一步降低了用户使用门槛。例如，尝试给出几个关键词，生成式人工智能能够在十几秒内轻松写出有人物、有事件、有观点的几百字新闻稿，大幅提升了宣传作品的供给速度。

(II) The “Empathy” Awareness: Precise and Intelligent Information Distribution

（二）“共情”意识：信息分发精准化与智能化

Generative artificial intelligence has highly intelligent characteristics, capable of generating and stimulating “empathy” awareness in the communication process, making propaganda content distribution more precise and effective. People interact instantly with intelligent tools in the “relay” training of repeated questioning and follow-up questioning, and obtain effective information. Unlike search engines, generative artificial intelligence exhibits performance closer to the human level in multiple fields, such as high-precision intelligent question-answering systems, advanced sentiment analysis, context-aware machine translation, and personalized intelligent recommendation systems. For example, in response to questions such as being in a bad mood or under great pressure, it can quickly generate answers such as emotional comfort, solutions, and detailed care. From the demand side of propaganda, ideology, and cultural work, generative artificial intelligence, relying on powerful data training and language models, can meet the personalized and differentiated needs of users, while relying on the analysis and mining of digital intelligence technology to provide valuable guidance and reference for propaganda, ideology, and cultural work. Generative artificial intelligence greatly promotes the precise distribution and docking of propaganda content, with low latency and high efficiency, saving the time cost of repeated searches, while filtering out complex and redundant information in the production and distribution process. By identifying implicit emotional tendencies in text and integrating multi-source knowledge such as users’ historical behavior and preferences, generative artificial intelligence highlights the effectiveness of docking the supply side and demand side of propaganda, ideology, and cultural work. In this process, it captures and analyzes the “instantaneous emotions” of user groups, tracks the dissemination patterns of propaganda ideological content, and thereby predicts group psychology and behavioral changes, further guiding ideological and political and ideological work, improving the effectiveness of public opinion monitoring and ideological guidance, making the efficiency of propaganda, ideology, and cultural work higher.

生成式人工智能具有高度智能化特征，能够在传播过程中产生并催发“共情”意识，使宣传内容分发更加精准化与效能化。人们在反复提问、追问的“接龙”训练中与智能工具即时互动，并获取有效信息。与搜索引擎不同，生成式人工智能在多领域展现出更为贴近人类水平的性能，如高精度智能问答系统、高级情感分析、上下文感知机器翻译和个性化智能推荐系统等应用实例。譬如，面对情绪不好、压力很大等提问能快速生成情感安慰、解决办法、细节关怀等回答。从宣传思想文化工作的需求侧看，生成式人工智能依托强大的数据训练和语言模型能够满足用户的个性化与差异化需求，同时依靠数智技术的分析和挖掘为宣传思想文化工作提供宝贵的指导和参考。生成式人工智能极大推动了宣传内容的精准分发对接，低延时、高效率，节约反复检索的时间成本，同时在生产与分发过程中过滤掉繁杂冗余的信息。生成式人工智能通过识别文本中的隐含情感倾向、整合用户的历史行为和偏好等多源知识，让宣传思想文化工作供给侧和需求侧对接的效能化凸显，在此过程中捕捉分析用户群体的“瞬间情绪”，追踪宣传思想内容的传播规律，进而预判群体心理和行为变化，以此进一步指导思想政治和意识形态工作，提升舆情监测和思想引导的效果，使得宣传思想文化工作的效能更高。

(III) The “Symbiosis” Scene: Panoramic and Three-Dimensional Communication Paradigm

（三）“共生”景象：传播范式全景化和立体化

Generative artificial intelligence accelerates the leap from “interpersonal communication” to “human-machine communication,” continuously extending the “tentacles” of propaganda, ideology, and cultural work, integrating into all aspects of work, study, interpersonal interaction, and lifestyle, presenting a “symbiosis” scene, and promoting the formation of a new communication paradigm of human-machine collaboration, spatiotemporal interaction, and cross-border sharing. The communication and technical capabilities of generative artificial intelligence will exhibit high interactivity and scalability, and will successively integrate more applications and fields. Therefore, broad-connection, multi-form, panoramic communication infinitely extends the tentacles of intelligently generated content, continuously enhancing human-machine interaction trust in the process of satisfying instant information. For users, the powerful information integration capability of generative artificial intelligence makes it a good helper in study and work, and its “humanized” logic and thinking ability also make it a powerful tool in life, and can form a sense of dependence on it in interactive communication. Artificial intelligence constructs an all-round information scene, and people achieve value identification through the perceptual experience and emotional resonance of “environment-body-consciousness.” It can thus be foreseen that generative artificial intelligence will gradually become an indispensable part of people’s study, work, cultural entertainment, and interactive communication, thereby more deeply empowering propaganda, ideology, and cultural work. Under the influence of artificial intelligence, all dimensions of individual and group growth and development will become the objects and starting points of propaganda, ideology, and cultural work, and make the work model more comprehensive and three-dimensional.

生成式人工智能加速了“人际传播”到“人机传播”的跨越，使宣传思想文化工作的“触角”不断延展，融入工作学习、人际交往、生活方式的全方位领域，呈现“共生”景象，助推形成人机协同、时空交互、跨界共享的传播新范式。生成式人工智能的传播能力和技术能力将表现出高度的交互性和扩展性，并陆续整合更多的应用和领域。因此，广连接、多形态、全景化的传播，使智能生成内容的触角无限延伸，在即时信息满足过程中不断增进人机交互信任。就用户而言，生成式人工智能强大的信息整合能力让其成为学习工作上的好帮手，“人性化”的逻辑和思考能力也让其成为生活中的得力工具，并能在互动交流中形成对其的依赖感。人工智能构造了全方位的信息场景，人们通过“环境－身体－意识”的感知体验和情感共振实现价值认同。由此可以预见，生成式人工智能将逐渐成为人们学习工作、文娱生活、交往互动中不可或缺的一部分，进而更深入地为宣传思想文化工作赋能。在人工智能影响下，个人和群体成长、发展的全维度都将成为宣传思想文化工作的对象和抓手，并使工作模式更加全程化和立体化。

(IV) The “Co-integration” Trend: Integration and Synergy of the Propaganda Field

（四）“共融”趋势：宣传场域一体化与协同化

At present, generative artificial intelligence at home and abroad is developing competitively, presenting a scene of “a hundred flowers blooming,” and the pace of multi-scenario application is constantly accelerating. In the context of the era of artificial intelligence, the globalized communication pattern dominated by intelligent communication, which transcends time and space, makes propaganda, ideology and cultural work increasingly show a trend of “integration,” including the integration and synergy among subjects, objects, media, and the environment. Taking the field of natural language processing as an example, future large models may possess cross-language, cross-domain adaptive and transfer learning capabilities. Generative artificial intelligence technology, based on data, computing power, and algorithms, shapes a new communication pattern and plays an important role in promoting the integration of the propaganda field. From the subject level, propaganda, cultural, and other departments can further integrate and jointly play a role in propaganda, ideology and cultural work, giving rise to the synergistic advantages of subject departments and personnel. From the object level, the target of the work is no longer a single information receiver but becomes part of the subject of propaganda, ideology and cultural work. In the communication field, individual subjective initiative is enhanced, and the distinction between subject and object becomes more blurred. From the media level, generative artificial intelligence breaks the limitations of traditional propaganda media, can integrate a variety of life and work application tools and resources, and carry out propaganda, ideology and cultural work in an immersive manner across different media. From the field level, generative artificial intelligence expands propaganda, ideology and cultural work to the entire field of economic, social, and cultural development, no longer limited to the dimensions of time and space, but extended to the whole field where individuals, groups, and society interact and influence each other. From the quality and efficiency level, generative artificial intelligence helps to create ultra-high-definition, ultra-high-speed information transmission channels, form precise information interaction, and build an integrated communication system, allowing the propaganda ecosystem to extend and expand infinitely.

当前，国内外生成式人工智能竞相发展，呈现“百花齐放”的景象，多场景应用进程不断加快。在人工智能时代语境下，以智能传播主导的超时空全球化的传播格局，使宣传思想文化工作也愈发呈现“共融”趋势，包括主体、客体、介质和环境之间的一体化和协同化。以自然语言处理领域为例，未来的大模型可能具备跨语言、跨领域的自适应和迁移学习能力。以数据、算力与算法为基础的生成式人工智能技术形塑着新的传播格局，在促进宣传场域一体化过程中有着重要作用。从主体层面来看，宣传、文化和其他部门能够进一步融合，共同在宣传思想文化工作中发挥作用，催生出主体部门与人员的协同优势。从客体层面来看，工作对象不再是单一的信息接收方，而是成为宣传思想文化工作主体的一部分，在传播场域中个体主观能动性增强，主、客体对象化区分更加式微。从介质层面来看，生成式人工智能打破了传统宣传媒介的局限性，能够把多种生活、工作应用工具和资源进行融合，在不同介质中浸润式地开展宣传思想文化工作。从场域来看，生成式人工智能则是把宣传思想文化工作扩大到整个经济社会文化发展场域中，不再局限于时间和空间维度，而是扩展到个体、群体与社会相互联系影响的全场域。从质效来看，生成式人工智能有助于打造超高清、超高速的信息传播通道，形成精准化的信息交互并构建一体化的传播体系，让宣传生态圈无限延伸、拓展。

II. Guarding Against Loopholes: Hidden Concerns of Generative Artificial Intelligence Empowering Propaganda, Ideology and Cultural Work

二、杜隙防微：生成式人工智能赋能宣传思想文化工作的隐忧

On the new journey of the New Era, the era of generative artificial intelligence has already arrived, with strategic opportunities and risks and challenges coexisting. Propaganda, ideology and cultural work faces new situations and new tasks. Since the release of ChatGPT by OpenAI at the end of 2022 to the release of DeepSeek’s R1 model in January 2025, generative artificial intelligence has shown explosive development, increasingly becoming an important front that cannot be ignored in propaganda, ideology and cultural work. Generative artificial intelligence has brought new engines to propaganda, ideology and cultural work, but its use has both advantages and disadvantages, requiring us to deeply grasp the development trends and laws, and to prevent and resolve potential risks.

新时代新征程，生成式人工智能时代已然来临，战略机遇和风险挑战并存，宣传思想文化工作面临新形势和新任务。自2022年年底OpenAI发布的ChatGPT到2025年1月DeepSeek的R1模型发布以来，生成式人工智能呈现井喷式发展，愈益成为宣传思想文化工作不可忽视的重要阵地。生成式人工智能为宣传思想文化工作带来了新引擎，但其使用是利弊并存的，需要我们深刻把握发展态势和规律，防范和化解潜在风险。

(I) The Vortex Risk of the Propaganda Environment: Implicit Penetration and Ideological Erosion

（一）宣传环境的旋涡风险：内隐渗透与意识形态侵蚀

The technological revolution represented by generative artificial intelligence technology is profoundly reshaping the pattern of ideological confrontation. The “innate” bias of generative artificial intelligence cannot be underestimated, and we must be wary of Western digital hegemonic tendencies. Generative artificial intelligence intensifies the competition among diverse trends of thought and ideologies, forming a triple-dimensional risk of “value-discourse-practice,” causing mainstream ideology to face risks and dilemmas of weak guiding power and dissolving trust.

以生成式人工智能技术为代表的科技革命，正在深刻重塑意识形态交锋的格局。生成式人工智能“天生”的倾向性不容小觑，要谨防西方数字霸权主义倾向。生成式人工智能加剧了多元思潮和意识形态竞争，将形成“价值－话语－实践”三重向度风险，让主流意识形态面临引导力不强、信任消解的风险和困境。

First, the high-speed, panoramic, and covert generation of content makes the propaganda environment more complex and uncertain, and propaganda work will face this “silent and fierce contest” in the ideological field to the greatest extent. Generative artificial intelligence technology itself is neutral, but the intelligently generated content provided to the audience has public opinion attributes or social mobilization capabilities. The inherent risks of data itself and its source channels, as well as the “human feedback” relied upon in the generation and optimization of algorithm models, will amplify deviations from mainstream value orientations. As a representative of the Western discourse system, ChatGPT uses significantly more English data, which will exacerbate ideological bias. Global competition in the field of large AI models has become increasingly intense. Compared with international English-language public datasets, there is still a large gap in high-quality domestic datasets, with insufficient data completeness and accuracy, and unsatisfactory model training results.

第一，高速、全景、隐蔽的内容生成使宣传环境变得更加复杂和不确定，宣传工作将最大程度地直面意识形态领域这场“无声且激烈的较量”。生成式人工智能技术本身是中立的，但面向受众提供的智能生成内容具有舆论属性或社会动员能力。数据本身和来源渠道等方面的原生风险，以及算法模型生成与调优过程中所依赖的“人类反馈”，都将放大偏离主流价值的导向。作为西方话语体系代表的人工智能，ChatGPT使用的英语数据明显偏多，会加剧意识形态偏见。人工智能大模型领域的全球竞争已趋白热化，相较于国际英文公开数据集，国内高质量数据集还存在较大差距，数据完整性、精确性不足，模型训练效果并不理想。

Second, the discourse subjects of traditional ideology are often concrete, and their discourse expression can create a “sense of distance.” Generative artificial intelligence forms ideological discourse through implicit “dialogue.” Its mode of speech is a question-and-answer process of continuous progression, supplementation, exploration, and excavation. The discourse style of “acknowledging shortcomings” and “not being condescending” adds more trust from users. For example, when faced with erroneous or unreasonable questions, DeepSeek will refuse to answer, which instead makes the dialogue interaction more “zero-distance.”

第二，传统意识形态的话语主体往往是具象的，话语表达会让人有“距离感”。生成式人工智能通过内隐式的“对话”形成意识形态话语，其言说方式是不断递进补充、探索挖掘的问答过程，“承认短板”和“不高高在上”的话语风格让用户平添了更多信任。譬如，在面对错误或不合理问题时，DeepSeek会拒绝回答，反而让对话互动更加“零距离”。

Third, behavioral practice is the realistic purpose of ideology and an important dimension of propaganda, ideology and cultural work. Generative artificial intelligence technology is changing with each passing day and will inevitably intervene more deeply in real situations. Young people will gradually cede their thinking, judgment, and practical abilities, leading to confusion in behavioral practice. “Vortex” is often used as a metaphor for a situation from which one cannot extricate oneself. As the mainstream users of digital technology and potential future promoters, Generation Z [2]2 A Chinese idiom from *The Book of Rites* (《礼记·中庸》), meaning that success depends on advance preparation. (referring to those born between 1995 and 2009, also known as the “Internet generation”) youth have immature and unstable thinking. In their dependence on the use of generative artificial intelligence, they are easily negatively influenced by it, thus becoming a key group for hostile forces to penetrate and divide. Youth are an important target for carrying out ideological work, and special vigilance is needed against ideological penetration and imperceptible implicit discipline in human-computer interaction, to avoid being drawn into the vortex of non-mainstream ideology.

第三，行为实践是意识形态的现实旨归，是宣传思想文化工作的重要向度。生成式人工智能技术日新月异，必将越来越深地介入现实情境，青年人会逐渐让渡思考力、判断力和实践力，出现行为实践的迷思。“旋涡”常被喻为使人不能自脱的境地。作为数字技术的主流用户和未来的潜在推动者，Z世代（指1995年—2009年间出生的人，也被称为“网络世代”）青年思想尚不成熟稳定，在对生成式人工智能形成使用依赖中很容易受其负面影响，因而成为敌对势力进行渗透分化的重点人群。青年是开展意识形态工作的重要对象，尤其要警惕人机交互中的意识形态渗透和不易察觉的内隐规训，避免卷入非主流意识形态的旋涡。

(II) The “Swamp” Risk of the Propaganda Front: Pseudo-Environment and Information Obscuration

（二）宣传阵地的“沼泽”风险：拟态环境与信息遮蔽

In the virtual environment constructed by generative artificial intelligence, the authenticity, objectivity, and comprehensiveness of original data are easily obscured, and it is easy to fall into information “swamps” such as content being taken out of context or cultural conflicts, thereby causing the propaganda front to fall into a “swamp” mode. Since its emergence, generative artificial intelligence has had a “hallucination” problem, and the prospect of purely eliminating false information through technical means is not optimistic.

在生成式人工智能构建的虚拟环境中，容易遮蔽原始数据的真实性、客观性、全面性，容易陷入内容被断章取义、发生文化冲突等信息“沼泽”，进而让宣传阵地陷入“沼泽”模式。生成式人工智能自出现以来就存在“幻觉”问题,纯粹通过技术手段消除虚假信息的前景并不乐观。

First, from the perspective of post-truthism, generative artificial intelligence shapes the context of post-truth, easily leading users to deal with facts based on the principle of “avoiding the important and dwelling on the trivial.” In this field, false information will further deepen structural problems, potentially becoming a channel for “toxic or even weaponized” information output, and the propaganda front can easily enter an information “turbulence” zone. Generative artificial intelligence can automatically generate content at high speed and in large quantities, and can also fill the communication space with false or misleading information in a short time. As a tool for communication across the entire field, it is easily maliciously exploited to form public opinion incidents. For example, in the recent Russia-Ukraine conflict, deep learning-based false propaganda and social bot public opinion control were widely used, making people feel as if they were in a “fog,” unable to see the “true face of Mount Lu.” [3]3 "The New Era" (新时代) refers to the period since the 19th National Congress of the CPC in 2017, characterized by Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era. Although generative artificial intelligence has good content generation capabilities, surpassing traditional artificial intelligence, the content fabricated by “deepfake technology” is often difficult to distinguish between true and false. False information of unknown origin can become a flashpoint of public opinion at any time, exacerbating the explosive and fission-like spread of online rumors. Faced with this situation, if users lack critical reflection on information, they will have no “ability to resist,” thereby affecting value judgments and behavioral habits. According to reports, multiple cases of using artificial intelligence technology to fabricate false and untrue information have occurred in China and have been cracked by relevant departments.

第一，从后真相主义视角看，生成式人工智能形塑了后真相的语境，容易使用户以“避重就轻”的原则来应对事实。在此场域中，虚假信息将进一步深化结构性难题，潜在地成为“有毒甚至武器化”信息输出的通道，宣传阵地容易进入信息“乱流”区。生成式人工智能可以高速、大量地自动生成内容，也能在短时间内让传播空间充斥着虚假信息或误导性信息，作为全场域传播工具极易被恶意利用而形成舆情事件。譬如，近年的俄乌冲突中，深度学习的虚假宣传、社交机器人的舆情控制等被广泛应用，使人们如置身“迷雾”，难识“庐山真面目”。尽管生成式人工智能具备良好的内容生成能力，超越了传统的人工智能，但“深伪技术”编出的内容往往真伪难辨，来源不明的虚假信息随时可能成为舆论爆点，加剧网络谣言的爆发式、裂变式传播。面对此情况，如果用户缺乏对信息的审思，那将毫无“招架之力”，进而影响价值判断和行为习惯。据报道，我国已出现多起利用人工智能技术炮制虚假不实信息的案件并被相关部门侦破的情况。

Second, from the perspective of content application and information governance, the information transmitted by generative AI in application scenarios is relatively privatized, achieving “personalized customization” of content. The original “gatekeeping” methods will fail, and propaganda positions can easily enter information “blind spots.” In algorithmic logic, the large language models of generative AI are built on a “black box” structure, making it impossible to definitively decompose their algorithmic logic. This results in a natural lack of fact-checking capability in public opinion information governance and the potential failure of traditional technical means such as keyword or topic blocking. The capacity for online public opinion and digital information governance urgently needs improvement to prevent propaganda positions from entering a quagmire. A “gatekeeping shift” from content to platform has become a practical necessity; deep participatory experience, panoramic screening of data and information, and the technical logic of language models are prerequisites for comprehensive oversight. In addition, vigilance is needed against the risk of users’ cognitive misplacement of “assistance” and “substitution” brought by generative AI, as well as the scientific ethics, legal, and moral issues it may trigger, such as content plagiarism and academic integrity. Generative AI deeply affects young people, who are a “susceptible” group, fostering a dependent trust in new technology and making them highly prone to risks. This will have a major impact on propaganda, ideology and cultural work.

第二，从内容应用和信息治理角度看，生成式人工智能在应用场景传递的信息相对隐私化，并实现内容的“私人订制”，原有的“把关”方式会失灵，宣传阵地易进入信息“盲区”。在算法逻辑中，生成式人工智能的大语言模型，其背后是“黑盒”结构，无法确切分解其算法逻辑，这就造成舆情信息治理时事实核查能力天然缺失、传统屏蔽关键词或话题的技术手段可能失效的后果。网络舆情和信息数字治理能力急需提升，避免宣传阵地进入沼泽区。从内容到平台的“把关转移”成为现实需要，深度的参与体验、数据信息的全景筛查和语言模型的技术逻辑成为全方位监管的前提。此外，还要警惕生成式人工智能给用户带来的“辅助”和“替代”错位认知风险，以及可能引发的内容剽窃、学术诚信等科学伦理、法律和道德问题。生成式人工智能深度影响作为“易感”人群的青年，使他们对新技术产生依赖式信任并极易产生风险，这将对宣传思想文化工作产生极大冲击。

(III) The Risk of Dilution of Propaganda Effectiveness: Communication Restructuring and Dialogue Segmentation

（三）宣传效能的稀释风险：传播重构与对话区隔

“Disembedding” refers to the “lifting out” of social relations from limited local contexts, allowing them to be reorganized across vast spans of time and space. [4]4 "People-centered" (以人民为中心) is a core principle of the CPC's governance philosophy, emphasizing development for the people. The internet makes disembedding possible. In the new communication field formed by generative AI, the communication process of human-machine “dialogue” will dissolve the trust mechanism of “copresence,” thereby triggering the risk of diluting propaganda effectiveness.

“脱域”是指社会关系从有限的地方性场景中“剥离出来”，从而能跨越广阔的时间－空间距离去重新组织社会关系。互联网为“脱域”提供可能，在生成式人工智能形成的新传播场域中，人机“对话”的传播过程将消解“在场”信任机制，进而引发宣传效能稀释的风险。

First, in the communication context of generative AI, the modes of “joint information production” and “information co-construction” create a situation of human-machine interaction and coexistence, one person one communication space, and a “ten thousand people, ten thousand faces” communication pattern, restructuring communication relations and the propaganda field. In the “dialogue” space of generative AI, the audience transforms into seekers, acquirers, and generators of information. The subject transmitting information may not be “human”; users become important nodes in information generation and dissemination, possessing more initiative and discursive power than ever before. Regardless of time or place, generative AI can imitate human language and emotions to converse with users. The cross-temporal and spatial interaction and satisfaction of needs make the propaganda field more convenient, personalized, and privatized. The “disembedded” chain of information generation, dissemination, and reception in generative AI is far removed from the environment of human control and understanding. Inaccuracies, distortions, and irregularities in this process will amplify the contradictions between the virtual and the real, exposing users to risks such as trust anxiety, cognitive tension, and behavioral dislocation.

第一，在生成式人工智能传播语境下，“联合信息生产”“信息共建”模态造就了人机交互共存、一人一传播空间、“万人万面”传播样态的局面，重构了传播关系和宣传场域。在生成式人工智能“对话”空间中，受众转变为信息的寻求者、获得者、生成者，信息传递主体未必是“人”，用户变成信息生成、传播的重要节点，比以往拥有更多主动权和话语权。无论何时何地，生成式人工智能均能模仿人类语言和感情与用户对话。跨时空的信息交互和需求满足，使得宣传场域变得更加便捷化、个性化和私密化。生成式人工智能“脱域式”的信息生成、传播、接收链条，远离了人类控制和理解的环境，其过程中的失准、失真、失范情况会放大虚拟与现实间的矛盾，让用户面临信任焦虑、认知紧张和行为错位等风险。

Second, the “cyberspace” constructed by generative AI obscures the “copresence” effectiveness of propaganda, ideology and cultural work. German media theorist Friedrich Kittler argued that media technology “no longer merely controls human muscles as in the past, but also takes over the functions of the human central nervous system.” [5]5 "Uphold fundamental principles and break new ground" (守正创新) is a key concept in Xi Jinping's cultural thought, combining adherence to core socialist values with innovation. “Disembedded companionship” restructures new social relationships. Beyond generating homework, papers, poetry, and images, generative AI can also serve as an object for people to confide in and as an intimate “tree hole,” enhancing intimacy and satisfaction in interaction. At the same time, due to a natural affinity for new technology, stereotypes of mainstream media, and the primacy effect in judging events, people are more likely to shift their attention from official media to intelligent media, thereby weakening existing propaganda effectiveness.

第二，生成式人工智能所构建的“赛博空间”隐化了宣传思想文化工作的“在场”效能。德国媒介理论家弗里德里希·基特勒认为，媒介技术“不再像过去那样只是控制人的肌肉，还接管了人的中枢神经系统的官能”。“脱域式陪伴”重构了新的社交关系，生成式人工智能除了可以生成作业、论文、诗歌、图片，还可以成为人们的倾诉对象和亲密“树洞”，增强互动中的亲密感、满足感。与此同时，出于对新技术的天然好感、对主流媒体的刻板印象和对事件判断的首因效应，人们更容易将关注度由官方媒体转向智能媒体，导致既有宣传效能被弱化。

Faced with the comprehensive and profound impact of intelligently generated content on people, the established model of propaganda, ideology and cultural work will be broken. If the existing discourse system of communication is not transformed in time, the appeal of offline propaganda activities and online communication platforms such as official websites, WeChat, and Weibo will decrease, leading to propaganda quality and effectiveness that cannot keep pace with the development of generative AI. Currently, generative AI is in a period of rapid development; its functions and potential have not yet been fully unleashed. In the future, it will bring both more opportunities and greater challenges to propaganda methods.

面对智能生成内容对人们的全面、深刻影响，宣传思想文化工作的固有模式将被打破，原有的传播话语体系如不及时转型，线下宣传活动以及线上官网、微信、微博等传播平台的吸引力会降低，将导致宣传质量和效果难以跟上生成式人工智能的发展脚步。目前，生成式人工智能处于高速发展时期，其功能和潜能尚未全部释放，未来将对宣传方式带来更多的机遇和提出更大的挑战。

(IV) The Risk of Reification of Propaganda Content: Cultural Constraints and the “Hollowing Out” Phenomenon

（四）宣传内容的物化风险：文化囿限与“空心”现象

In the era of digital globalization, the vigorous development of Western AI technology has enabled consumerist culture to spread further, presenting a trend of cultural centralization and squeezing the space for the inheritance and development of fine traditional Chinese culture. In the New Era, the role of culture in invigorating the national spirit, maintaining national identity, and promoting the comprehensive development of individuals is increasingly prominent. Cultural construction is an important part of propaganda, ideology and cultural work. From the perspective of the inheritance and development of fine traditional Chinese culture, over-reliance on AI technology may lead to the reification of propaganda content and the “hollowing out” limitation of superficial and formalistic cultural construction, thereby trapping cultural construction in a predicament.

在数字全球化时代，西方人工智能技术的蓬勃发展使消费主义文化得以进一步蔓延，呈现文化中心化趋势，对中华优秀传统文化的传承与发展空间形成挤压。新时代新征程，文化在振奋民族精神、维系国家认同、促进人的全面发展等方面的作用日益凸显，文化建设是宣传思想文化工作中的重要一环。从中华优秀传统文化的传承发展看，如果过于依赖人工智能技术可能会导致宣传内容物化以及文化建设表面化、形式化的“空心”局限，进而使文化建设囿于困局。

First, from the perspective of the generation and supply of fine traditional cultural content, generative AI dominated by Western values is clearly characterized by “valuing material culture while neglecting spiritual culture.” Generative AI is not omniscient; it relies on a vast database of texts and requires training to form effective language models. The more advanced and powerful the algorithm, the more it depends on the precision and breadth of the training data. It should be noted that AI tools are clearly incapable of original knowledge discovery; they can only provide relatively logical answers to general questions after possessing massive amounts of data. From data to algorithms, they lack core cultural “understanding” and “interpretive power.” Furthermore, when processing data on the history and culture of different countries or regions, generative AI often exhibits imbalances in data volume, with insufficient supply of cultural content for some countries.

第一，从优秀传统文化内容生成和供给来看，西方价值观主导的生成式人工智能明显带有“重物质文化、轻精神文化”的色彩。生成式人工智能并不是全知全能的，它要依靠基数庞大的文本数据库，还要对其加以训练形成有效的语言模型。越是先进强大的算法，越要依赖训练数据的精度和广度。需要注意的是，人工智能工具显然无法做到原创性的知识发现，仅能在拥有海量数据后，对一般问题给出比较有逻辑的解答，从数据到算法都缺乏核心的文化“理解力”和“解释力”。此外，生成式人工智能在处理不同国家或地区的历史文化方面的数据时，往往存在载量不同的情况，对一些国别的文化内容供给远远不够。

Second, from the perspective of the mechanism of cultural constraints, the imbalance in the supply of cultural content can lead to a blunting of people’s perception in the empty and tedious AI cultural environment. Baudrillard argued that “consumer society is also a society of consumer training, of social domestication oriented toward consumption.” [6]6 The National Conference on Propaganda, Ideology and Cultural Work, held in October 2023, was a high-level meeting that formally proposed Xi Jinping Thought on Culture. A large amount of intelligently generated content—knowledge-based, common-sense, and entertainment-oriented—gains significant attention and high user stickiness, directly leading to a structural imbalance in the output of propaganda content supply. Amid the vast array of diverse digital information, if the supply of fine traditional Chinese cultural works is insufficient, users will develop a tendency toward cultural perceptual blunting and reification. Perhaps textual work can be replaced by intelligent generation, but the profound and extensive fine traditional culture of the Chinese nation can never be replaced. This is the core key to propaganda, ideology and cultural work, and it is also the pain point and critical point that cultural construction urgently needs to address.

第二，从文化囿限的影响机理来看，文化内容供给的失衡会造成人们在空虚乏味的人工智能文化环境中出现感知力钝化情形。波德里亚认为，“消费社会也是进行消费培训、进行面向消费的社会驯化的社会”。大量知识性、常识性、娱乐性等智能生成内容获得较多的关注度和较高的用户黏性，直接导致宣传内容供给产出的结构失衡。在纷繁多样的数字信息中，如果中华优秀传统文化作品供给不足，就会让广大用户产生文化感知钝化和物化的倾向。也许文字工作可以被智能生成替代，但是中华民族博大精深的优秀传统文化却是永远无法被替代的，这是宣传思想文化工作的核心密钥，更是文化建设亟须重视的痛点和关键点。

Young people are important inheritors, practitioners, and disseminators of fine traditional Chinese culture. For propaganda work, accelerating and increasing the supply of high-quality digital cultural works to better meet the cultural life needs of youth groups is an urgent necessity. The momentum of AI development is unstoppable. Promoting the deep integration of the traditional advantages of propaganda, ideology and cultural work with technological development, and ensuring that high-quality cultural works maintain their endogenous power and appeal, are important tasks currently facing us.

青年人是中华优秀传统文化的重要传承者、践行者和传播者，对宣传工作来说，加快、加大优质数字文化作品供给，更好满足青年群体文化生活需求迫在眉睫。人工智能的发展势头不可阻挡，推动宣传思想文化工作传统优势同技术发展高度融合，让优质文化作品保持内生力、感染力是当前面临的重要课题。

III. Breaking Out of the Cocoon: Innovative Paths for Propaganda, Ideology and Cultural Work in the Era of Generative AI

三、破圈而出：生成式人工智能时代宣传思想文化工作的创新路径

Propaganda, ideology and cultural work must provide a strong ideological guarantee, a powerful spiritual force, and favorable cultural conditions for comprehensively building a modern socialist country and comprehensively advancing the great rejuvenation of the Chinese nation. Facing the tide of the generative AI era, it is necessary to reflect on and analyze traditional working methods, integrate digital-intelligent thinking throughout the entire process of propaganda, ideology and cultural work, directly confront the new battlefield of AI and digital space, truly seize the opportunity by moving with the trend, fully advance into the main battlefield, turn “variables” into “increments,” and firmly occupy the commanding heights of communication in guiding public opinion, leading thought, inheriting culture, and serving the people.

宣传思想文化工作要为全面建设社会主义现代化国家、全面推进中华民族伟大复兴提供坚强思想保证、强大精神力量、有利文化条件。面对生成式人工智能的时代大潮，要反思和辨析传统工作方法，将数智化思维贯穿于宣传思想文化工作的全过程，直面人工智能与数字空间新战场，真正做到迎势出击、抢占先机，全面挺进主战场，化“变量”为“增量”，牢牢占据舆论引导、思想引领、文化传承、服务人民的传播制高点。

(I) Consolidating the Foundation: Disenchanting Technology, Strengthening Value Guidance, and Maintaining Political Determination

（一）固本：技术祛魅，强化价值引领，保持政治定力

The disenchantment of technology must achieve the dissolution of the mystery, sacredness and allure of science and knowledge, that is, to reveal the role generative artificial intelligence plays in propaganda, ideology and cultural work as well as the essential laws of its development and evolution. Political guidance is its directional goal. Propaganda, ideology and cultural work in the era of new technology must firmly hold the political stance, avoiding "the leadership of the human-machine system being handed over to the machine." [1]1 A reference to the concern that in human-machine systems, control may shift from humans to machines, a theme in discussions of technological autonomy. Scientific and technological innovation is permeated with value relations, and artificial intelligence is no exception. The development of generative artificial intelligence must adhere to correct value orientation in order to move toward a powerful and beautiful future.

技术祛魅要做到对科学和知识的神秘性、神圣性、魅惑力的消解，即揭示生成式人工智能在宣传思想文化工作中发挥的作用以及发展演进的本质规律。政治引领是其目标指向。新技术时代的宣传思想文化工作必须守牢政治立场，避免“人机系统的领导转交给了机器”。科技创新渗透着价值关系，人工智能也不例外。生成式人工智能的发展必须坚持正确的价值导向，才能走向强大而美好的未来。

The explosive popularity of generative artificial intelligence indicates a clear breakthrough in AI technology, behind which lies the development logic of "the strong become stronger" in the future field of AI production. The essence of a generative AI platform is the concretization of a digital bilateral market, and its formation and operational mechanism are also based on digital communication. Preparedness ensures success; [2]2 A Chinese idiom from *The Book of Rites* (《礼记·中庸》), meaning that success depends on advance preparation. in the technological development and scenario application of generative AI, value guidance must be strengthened in order to maintain the political steadfastness and developmental vitality of propaganda, ideology and cultural work.

生成式人工智能的爆火表明人工智能技术的明显突破，其背后是未来人工智能生产领域“强者更强”的发展逻辑。生成式人工智能平台的本质是数字化双边市场的具象化，其形成与运行机制也是基于数字传播。凡事预则立，在生成式人工智能的技术发展和场景应用中，必须强化价值引领，才能保持宣传思想文化工作的政治定力、发展活力。

From the perspective of post-humanism, generative artificial intelligence, as a cognitive tool, embodies a certain value implication in AI technologies such as databases and intelligent algorithms. On the one hand, as users of generative AI, faced with the super-strong production capacity of intelligently generated content, propaganda work must adhere to political security as the foundation, leverage the advantages of propaganda and education, and be adept at using mainstream values to guide users in mastering the "algorithm," truly making it serve our purposes. On the other hand, as participants in the technological R&D and innovation of generative AI, relevant entities must leverage their disciplinary, intellectual and talent advantages to ensure the legality and reasonableness of pre-training data sources, while simultaneously strengthening innovation in technologies such as intelligent algorithms, data training and language models, accelerating the pace on the global AI track, consolidating the dominant position of mainstream ideology, and further enhancing international discourse power and leading influence. In practice, only through the coordinated advancement of value guidance, moral ethics and technological application as a "trinity" can we better create, embrace and respond to the "new epoch" of AI-generated content.

从后人类主义视角看，生成式人工智能作为一种认知工具，数据库和智能算法等人工智能技术也体现一定的价值意蕴。一方面，作为生成式人工智能的应用者，面对智能生成内容的超强生产能力，宣传工作必须坚持政治安全为根本，要发挥宣传教育优势，善于用主流价值引导用户驾驭“算法”，真正将其为我所用。另一方面，作为生成式人工智能的技术研发和创新参与者，相关主体要发挥学科、智力、人才优势，保障预训练数据来源的合理合法，同时加强智能算法、数据训练和语言模型等技术方面创新，在全球人工智能赛道跑出加速度，巩固主流意识形态主体地位，进一步增强国际话语权和主导力。在实践过程中，只有价值引领、道德伦理与技术应用“三位一体”的协同推进，才能更好地开创、拥抱和应对人工智能生产内容“新纪元”。

(II) Cultivating the Foundation: Information Noise Reduction, Strengthening Mainstream Public Opinion, Activating Positional Effectiveness

（二）培元：信息降噪，壮大主流舆论，激活阵地效能

Consolidate and strengthen the mainstream ideological and public opinion of forging ahead in the New Era, [3]3 "The New Era" (新时代) refers to the period since the 19th National Congress of the CPC in 2017, characterized by Xi Jinping Thought on Socialism with Chinese Characteristics for a New Era. focus on strengthening positive publicity with strong confidence as the key, and improve the ability to guide public opinion. In the era of generative AI, the internet is complex and volatile, and communication technologies, methods and means are more diverse. It is necessary to change thinking in terms of content supply, literacy cultivation, management and governance, further strengthen the management and construction of propaganda positions, and expand the mainstream public opinion ecosystem.

巩固壮大奋进新时代的主流思想舆论，以强信心为重点加强正面宣传，提高舆论引导能力。生成式人工智能时代，网络复杂多变，传播技术、方法和手段更加多样，需要从内容供给、素养培育、管理治理等方面转变思路，进一步增强宣传阵地管理建设，壮大主流舆论生态圈。

First, grasp the initiative with strategic thinking and use content to precipitate the mainstream. Integrate AI technology with a people-centered [4]4 "People-centered" (以人民为中心) is a core principle of the CPC's governance philosophy, emphasizing development for the people. approach organically, look to the future, follow the general trend, and let generative AI empower the propaganda field. Ride the express train of new technology, seize the new intelligent track, accelerate content supply, enrich high-quality content, gather resources and strength, to enhance propaganda effectiveness and gain greater advantages. Focus on the thoughts and needs of target audience groups, strengthen and expand online propaganda that warms hearts and strengthens confidence, create distinctive brands, enrich the lives of young people, and make mainstream content the "true mainstream" in cyberspace. Second, uphold fundamental principles and break new ground [5]5 "Uphold fundamental principles and break new ground" (守正创新) is a key concept in Xi Jinping's cultural thought, combining adherence to core socialist values with innovation. with systematic thinking, and use education to purify technology. The authenticity of intelligently generated content is the touchstone for the healthy survival and development of generative AI. Faced with vast amounts of complex information, there must be a process of "clarification" to achieve a state of unconcealment and dispel the allure of information. Advance the cultivation of AI literacy, strengthen users' judgment and comprehension of online public opinion and generated information, especially the ability to prudently discern intelligent content and to question, evaluate and select information, and fortify the ideological defense line of the online position. Third, hold the bottom line with rule-of-law thinking, and improve the ability to manage and govern according to law. In August 2023, China's first national-level regulatory document for generative AI, the "Interim Measures for the Management of Generative Artificial Intelligence Services," came into effect, which states that "the state adheres to the principle of balancing development and security, and combining the promotion of innovation with law-based governance, and takes effective measures to encourage the innovative development of generative artificial intelligence," providing guidance for managing and using new AI technologies well. In March 2025, the Cyberspace Administration of China issued the "Measures for the Identification of AI-Generated Synthetic Content." Propaganda, ideology and cultural work must further guard the security line, enhance the ability for "human-machine cooperative governance" and build an online information emergency mechanism, to further consolidate the online position. The propaganda, ideology and cultural front must coordinate forces such as technology and education to effectively apply new concepts, new technologies and new means, improve service capabilities and governance levels in the process of R&D and use of generative AI, and strengthen propaganda, education and intelligent gatekeeping in ideological, technical and legal aspects.

一是以战略思维把握主动，用内容沉淀主流。要将人工智能技术与以人为本有机融合起来，放眼未来、顺应大势，让生成式人工智能赋能宣传领域。要搭乘新技术的快车，抢占智能新赛道，加快内容供给、丰富优质内容、汇聚资源力量，以提升宣传效能、赢得更大优势。要聚焦受众群体所想所需，做强做大暖人心、强信心的网络宣传，办出特色品牌，丰富青年生活，让主流内容在网络空间成为“真主流”。二是以系统思维守正创新，用教育净化技术。智能生成内容的真实与否是生成式人工智能得以健康生存、发展的试金石。面对大量庞杂的信息，要有“澄明”的过程，以达到无遮蔽状态，祛除信息魅惑。要前置人工智能素养培育，加强用户对网络舆论、生成信息的判断力和领悟力，尤其是具备审慎明辨智能内容以及质疑、评估和选择信息的能力，筑牢网络阵地的思想防线。三是以法治思维守牢底线，提高依法管理治理能力。2023年8月，我国首个国家级生成式人工智能监管文件《生成式人工智能服务管理暂行办法》施行，其中指出“国家坚持发展和安全并重、促进创新和依法治理相结合的原则，采取有效措施鼓励生成式人工智能创新发展”,为管好用好人工智能新技术提供了指南。2025年3月，国家网信办发布了《人工智能生成合成内容标识办法》。宣传思想文化工作更要守牢安全防线，增强“人机合作治理”能力和建设网络信息应急机制，进一步巩固网络阵地。宣传思想文化战线要协同技术、教育等力量来有效运用新理念、新技术和新手段，提高在研发、使用生成式人工智能过程中的服务能力和治理水平，加强思想、技术和法律方面的宣传教育和智慧把关。

(III) Forging the Soul: Supplementing Information Sources, Enriching Cultural Supply, Cultivating Endogenous Momentum

（三）铸魂：信源补缺，丰盈文化供给，厚植内生动力

In October 2023, the National Conference on Propaganda, Ideology and Cultural Work [6]6 The National Conference on Propaganda, Ideology and Cultural Work, held in October 2023, was a high-level meeting that formally proposed Xi Jinping Thought on Culture. first proposed Xi Jinping Thought on Culture, [7]7 Xi Jinping Thought on Culture (习近平文化思想) is the latest addition to the theoretical system of Xi Jinping Thought, focusing on cultural development and confidence. providing a powerful ideological weapon and scientific action guide for doing a good job in propaganda, ideology and cultural work on the new journey of the New Era and shouldering new cultural missions. "For cultural construction, upholding fundamental principles prevents losing oneself and losing direction, while innovation enables grasping the times and leading the times." [8]8 A quote from Xi Jinping, emphasizing the dialectic between preserving core principles and pursuing innovation in cultural work. It is necessary to grasp culture as the driving force, let generative AI empower fine traditional Chinese culture, enrich cultural supply, strengthen cultural confidence, and cultivate the endogenous momentum of propaganda, ideology and cultural work. With generative AI, users from different ethnic groups around the world can transcend language barriers to "see a bigger world." In March 2024, China Global Television Network (CGTN) launched "The Legend of Nezha: Heroes Born" for overseas youth, using generative AI to adapt the script, dub and score, and shape character images, showcasing the family-country sentiment and sense of responsibility in traditional Chinese values to the world. From the perspective of disseminating fine traditional Chinese culture, generative AI is expected to shape an integrated international communication field and international cultural exchange platform, while broadening the dissemination field of fine traditional Chinese culture, it also helps enhance China's cultural soft power.

2023年10月，全国宣传思想文化工作会议首次提出习近平文化思想，为做好新时代新征程宣传思想文化工作、担负起新的文化使命提供了强大思想武器和科学行动指南。“对文化建设来说，守正才能不迷失自我、不迷失方向，创新才能把握时代、引领时代。”要抓好文化这一牵引，让生成式人工智能赋能中华优秀传统文化，丰盈文化供给，坚定文化自信，厚植宣传思想文化工作的内生动力。有了生成式人工智能，来自世界不同民族的使用者能跨越语言障碍“看见更大的世界”。2024年3月，中国国际电视台（CGTN）面向海外青年推出《自古英雄出少年之哪吒传奇》，用生成式人工智能改编剧本、配音配乐、塑造人物形象，向全球展现了中华传统价值观中的家国情怀和责任担当。从中华优秀传统文化传播角度看，生成式人工智能有望形塑一体化的国际传播场域和国际文化交流平台，在拓宽中华优秀传统文化传播场域的同时，也有助于提升我国的文化软实力。

First, build and improve a database of fine traditional Chinese culture with "Chinese flavor," create a large language model with Chinese characteristics, provide more sufficient and accurate knowledge information and language systems for generative AI, make the entire chain from data sources, intelligent algorithms, content generation to dissemination more "culturally rich," realize the technical expression of traditional culture, and give technology more cultural value. For example, the Natural Language Processing and Social Humanities Computing Research Center of the Department of Computer Science and Technology at Tsinghua University has developed the Chinese poetry generation system "Jiuge." [9]9 "Jiuge" (九歌) is an AI poetry generation system named after a classical Chinese poetic work, developed at Tsinghua University. Second, use AI technology to analyze the trending preferences and individual needs of user groups, conduct more precise propaganda and promotion, while effectively combining with intelligent media to complete deep learning, enhance the interactivity and interest of studying fine traditional Chinese culture, and explore secondary innovation of fine traditional Chinese culture. Third, in the context of cultural propaganda influenced by new technologies, understand human-computer interaction models, read language processing models, explore human-machine collaborative application scenarios, and do a good job in creating and disseminating high-quality cultural works for youth groups in the era of intelligent media. In the future, it is necessary to leverage the advantages of technological innovation and cultural education, assist in the creative transformation and innovative development of fine traditional Chinese culture in the AI era, let Chinese civilization shine brilliantly, and go abroad and internationalize with greater confidence.

一是构建和完善具有“中国味道”的中华优秀传统文化数据库，打造具有中国特色的大语言模型，为生成式人工智能提供更充分、更精准的知识信息和语言体系，让数据源头、智能算法、内容生成和传播等全链条更有“文化气息”，实现传统文化的技术表达，让技术更有文化价值。比如，清华大学计算机科学与技术系自然语言处理与社会人文计算研究中心就研发了中文诗歌生成系统“九歌”。二是借助人工智能技术分析用户群体的趋向爱好和个体需求，更精准地进行宣传推广，同时有效地与智媒相结合完成深度学习，增强中华优秀传统文化研学的互动性、趣味性，并探索中华优秀传统文化的二次创新。三是要在新技术影响下的文化宣传语境中，理解人机交互模式、读懂语言处理模型、探索人机协同应用场景，做好智媒时代面向青年群体优质文化作品的创作和传播。未来，要发挥科技创新、文化育人优势，助力人工智能时代中华优秀传统文化的创造性转化和创新性发展，让中华文明大放异彩，更自信地走出国门、走向国际。

Understand the substance and master its function, integrate substance and function, [10]10 "Understand the substance and master its function, integrate substance and function" (明体达用、体用贯通) is a classical Chinese philosophical concept, here applied to cultural work, meaning to grasp the essence and apply it effectively. focus on the new cultural mission, propaganda, ideology and cultural work must take the initiative and will have great potential. Data from the Bilibili video website shows that each month, over 80 million users watch content related to generative AI on the platform. The sticky users of generative AI are more open and sensitive to new technologies, and while participating in AI innovation, they can also fully experience the charm and appeal of fine traditional Chinese culture. It is suggested that propaganda and cultural departments fully leverage the important role of AI technology in cultural innovation and inheritance, making fine traditional Chinese culture "come alive" and gain more "fans."

明体达用、体用贯通，聚焦新的文化使命，宣传思想文化工作要主动作为，也将大有可为。哔哩哔哩视频网站数据显示，每个月有超过8000万用户在该平台观看生成式人工智能相关的内容。生成式人工智能的黏性用户对新技术更加开放、敏锐，在参与人工智能创新事业中也能充分感受中华优秀传统文化的魅力和感召力。建议宣传文化等部门充分发挥人工智能技术在文化创新与传承中的重要助推作用，让中华优秀传统文化“活”起来、“粉丝”多起来。

(4) Educating People: Elevating Dialogue, Strengthening the Work Force, and Advancing the Transformation and Upgrading of Discourse

（四）育人：对话升维，建强工作队伍，推进话语升级转型

In the era of generative artificial intelligence, the space for dialogue in propaganda, ideology and cultural-ethical work is continuously expanding and changing. The binary “real-world field + online field” is interconnected and its boundaries are merging, moving further toward all-weather, all-time, and all-domain coverage. The discourse system is the carrier for disseminating ideas and the media content of the online propaganda, ideology and cultural-ethical front. It is necessary to strengthen the construction of the work force, enhance the effectiveness of online ideological and political education, and advance the transformation and upgrading of the discourse of propaganda, ideology and cultural-ethical work in the era of generative artificial intelligence. To guard against the potential risks of generative AI, in addition to relying on policies and platform forces, propaganda and education are indispensable.

在生成式人工智能时代，宣传思想文化工作的对话空间不断拓展更迭，二元“现实场域+网络场域”互联互通、边界共融，更加迈向全天候、全时段、全场域。话语体系是传播思想的载体，是网络宣传思想文化阵地的媒介内容。要强化工作队伍建设，提升网络思想政治教育成效，推进生成式人工智能时代宣传思想文化工作话语升级转型。防范生成式人工智能的潜在风险，除了依靠政策和平台力量，宣传教育必不可少。

First, the subjects of propaganda, ideology and cultural-ethical construction must keep pace with the times, proactively “act first,” and enhance the appeal and “freshness” of discourse expression methods. Deep learning models make communication segmented; it is necessary to promote the integration of propaganda content, media resources, and talent force, so that discourse keeps pace with the times, increases discourse weight, expands discourse reach, strengthens precision communication, and expands the territory of mainstream value discourse; change the previous one-way output and grand narrative discourse, avoid the propaganda, ideology and cultural-ethical work falling into a state of “aphasia” or “silence,” and be adept at using technology to explain and enliven discourse, grasp the law of “change” and “constancy” in propaganda, and enhance the dissemination effectiveness of propaganda discourse. Second, propaganda, ideology and cultural-ethical workers must leverage their “presence” advantage, be adept at “optimizing action,” grasp the actual needs of propaganda targets in the era of artificial intelligence, and make discourse targeted and fresh. The obscuring nature brought by generative AI conceals the real identity of users; the immediacy, personalization, and transcendence of time and space of intelligently generated content greatly weaken people’s semantic understanding, emotional identification, and other aspects. It is necessary to connect the online and the real, respond to emotions, feelings, and actual needs, and speak discourse in a way that resonates, so as to improve the quality and efficiency of propaganda work. Third, it is especially necessary to grasp the core of discourse content, dare to act, pool strength to enhance guiding power, grasp the orientation of discourse from the perspective of values and cognition, and strengthen digital literacy for good communication with intelligent tools. In December 2024, the Ministry of Education issued a notice to deploy and strengthen AI education in primary and secondary schools, emphasizing the need to guide students to correctly handle the relationship between humans, technology, and society, and to promote large-scale teacher supply, etc. [11]11 The Ministry of Education notice refers to a December 2024 document titled “Notice on Strengthening Artificial Intelligence Education in Primary and Secondary Schools,” which calls for integrating AI literacy into curricula and training teachers. Generative AI has changed the traditional “speaking” process of propaganda and education. Discourse subjects must focus on discourse innovation, concentrate on enhancing people’s innovative ability, cognitive ability, and critical thinking, and achieve diversified and precise propaganda and education. The effective guidance of propaganda, ideology and cultural-ethical work must be realized through the transformation and upgrading of discourse, and efforts must be made to explore innovative paths for educating people in the digital intelligence era.

一是宣传思想文化建设主体要应时而上，主动“先为”，增强话语表达方式的吸引力和“鲜活度”。深度学习模型让传播分众化，要推进宣传内容、媒介资源、人才队伍力量整合，让话语与时俱进，提升话语权重、拓展话语辐射、强化精准传播，扩大主流价值话语版图;改变以往单向输出、宏大叙事的话语，避免宣传思想文化工作处于“失语”“失声”状态，要善用技术讲深讲活话语，把握好宣传“变”与“不变”的规律，增强宣传话语传播效能。二是宣传思想文化工作者要发挥“在场”优势，善于“优为”，把握好人工智能时代宣传对象的现实需求，让话语对象化、新鲜化。生成式人工智能所带来的遮蔽性隐去了用户的现实身份，智能生成内容的即时性、个性化、超时空化使人们在语义理解、情感认同等方面大幅减弱，要将网络与现实贯通，回应情绪、情感和实际需要，将话语讲到心坎上，才能让宣传工作提质增效。三是尤其要把握话语内容的核心，敢于作为，汇聚合力增强引导力，站在价值与认知层面把握好话语导向，强化与智能工具进行良好沟通的数字素养。2024年12月，教育部印发通知部署加强中小学人工智能教育，强调要引导学生正确处理人与技术、社会的关系，推动规模化教师供给等。生成式人工智能改变了传统宣传教育“言说”过程，话语主体要注重话语的创新，聚焦增强人的创新能力、认知能力和批判性思维，实现多样化、精准化宣传教育。宣传思想文化工作的有效引领要通过话语升级转型而实现，要着力探索数智时代育人创新路径。

In summary, propaganda, ideology and cultural-ethical work is a systematic project with a long way to go. Generative AI accelerates the trend of cultural diversification and social networking. Development opportunities coexist with difficulties and challenges. It must be aligned with the new trends, new requirements, and new developments of the New Era, use the power of new technologies to resolve new contradictions and new risks, and display new phenomena and new achievements. At the cusp of the intelligent revolution, facing a population with vibrant ideas and innovative potential, propaganda, ideology and cultural-ethical work can only truly “break out of the circle” by continuing to move toward innovation, collaboratively advancing the development of ideology and public opinion, civilization and culture, internal and external propaganda, and online and offline spheres, and broadly consolidating ideas and strength.

总之，宣传思想文化工作是一项系统工程，任重道远。生成式人工智能加速了文化多样化、社会网络化的发展大势，发展机遇与困难挑战并存，必须与新时代新趋向新发展接轨、以新技术之力化解新矛盾新风险，展现新气象新作为。处在智能革命的风口，面向思想有活力、创新有潜力的人群，宣传思想文化工作唯有继续迈向创新，协同推进思想舆论、文明文化、内宣外宣和网上网下发展，广泛凝聚思想和力量，才能真正“破圈而出”。

(The author is a lecturer in the Propaganda Department of the Party Committee of Donghua University.)

（作者系东华大学党委宣传部讲师）

Online editor: Ma Jingren

网络编辑：马京仁