Ethical Risks of Generative AI Empowering Ideological and Political Education and Pathways for Resolution — An Examination from the Perspective of Educated Subjects’ Subjectivity

At present, the widespread application of generative artificial intelligence, represented by large language models, has greatly enriched and expanded the practical scenarios of “AI + education,” driving the traditional educational ecology to continuously evolve in a digital and intelligent direction. In recent years, China has deeply implemented its education digitalization strategy, actively exploring the deep integration of intelligent technology with educational practice. Top-level designs and policy documents for education digitalization and intelligentization have been continuously introduced and implemented in an orderly manner. The inevitability of technology-mediated existence in education is rapidly transforming education into a new form that blends the virtual and the real. Empowered by generative AI, ideological and political education has gained powerful transformative force. Leveraging comprehensive technological advantages such as natural language processing, big data analysis, deep learning, and multimodal integration, generative AI has expanded the space for innovation in ideological and political education in areas such as personalized learning, knowledge graph construction, and immersive interaction, thereby exerting a significant impact on the value shaping, knowledge acquisition, and ability cultivation of educated subjects. However, “technology can be a sharp tool for development, but it may also become a source of risk.” The one-dimensional expansion of instrumental rationality in intelligent technology can also dissolve the foundation of value rationality in ideological and political education, causing a deep “colonization” of the educational process by technological logic. This induces a stress-induced alienation of the educated subject’s subjectivity and triggers a series of ethical risks, urgently requiring the active exploration of countermeasures from both theoretical and practical levels.

当前，以大语言模型为代表的生成式人工智能的广泛应用，极大地丰富和拓展了“人工智能+教育”的实践场景，促使传统教育生态不断向数字化和智能化方向演进。近年来，我国深入实施教育数字化战略，积极探索智能技术与教育实践的深度融合，教育数字化、智能化的顶层设计和政策文件不断出台并有序落地。教育技术化生存的必然性促使教育快速向虚实融合的新样态转变，思想政治教育在生成式人工智能的赋能下获得强大的变革性力量。生成式人工智能凭借自然语言处理、大数据分析、深度学习、多模态融合等综合性技术优势，拓宽了思想政治教育在个性化学习、知识图谱构建和沉浸式交互等方面的创新空间，进而对受教育者的价值塑造、知识习得和能力培养产生重要影响。但是，“科技是发展的利器，也可能成为风险的源头。”智能技术工具理性的单向度扩张也会消解思想政治教育的价值理性根基，造成技术逻辑对教育过程的深度“殖民化”，促使教育对象的主体性发生应激性异化并诱发一系列伦理风险，亟须从理论和现实层面积极探寻应对之策。

I. The Technological Logic by Which Generative AI Empowering Ideological and Political Education Leads to the Dissolution of Subjectivity

一、生成式人工智能赋能思想政治教育引致主体性消解的技术逻辑

Marx pointed out: “The species-character of the human being is precisely free, conscious activity.” The cultivation of the educated subject’s subjectivity in ideological and political education is an important pathway for realizing the free and comprehensive development of the human being. Subjectivity is manifested in practice across multiple dimensions, including autonomy, self-activity, initiative, consciousness, and freedom. The contradiction between generative AI technology and ideological and political education is largely embedded in technology’s function of dissolving subjectivity. By constructing a technological explanatory framework based on data-driven discipline, algorithmic cognitive substitution, and technological mediation, it is possible to provide the necessary theoretical fulcrum for deconstructing the ethical risks of ideological and political education in the digital-intelligence era.

马克思指出：“人的类特性恰恰就是自由的自觉的活动”。思想政治教育对教育对象主体性的培养，是实现人的自由全面发展的重要途径。主体性在实践中体现在多个维度，包括自主性、自为性、能动性、自觉性、自由性等方面。生成式人工智能技术与思想政治教育之间的矛盾，很大程度上内嵌于技术对主体性的消解功能之中，而通过构建基于数据化规训、算法认知替代、技术中介化的技术解释框架，能够为解构数智时代思想政治教育的伦理风险提供必要的理论支点。

(I) Data-Driven Discipline Leads to the Technological Deconstruction of Value Subjectivity

（一）数据化规训引发价值主体性的技术性解构

Values education is an important pathway for maintaining the subjective status of the educated subject and achieving identification with mainstream ideology. Through an interactive mechanism based on intersubjectivity, it forms a value consensus in the subject’s value cognition, judgment, and selection, thereby achieving the goal of cultivating value rationality. At the philosophical-ontological level, value subjectivity links objective value with the subject’s existence. The diachronic evolution of value forms always revolves around the historical existence of the subject, thereby constructing a transformation mechanism of “subject awakening — value identification — practical consciousness” at the practical dimension of education, enabling individual value choices to remain highly consistent with the guiding core social values. However, in the context of generative AI’s deep involvement in ideological and political education, data-driven discipline gradually forms a new mode of power operation by reshaping the value structure of the educated subject. Utilizing technical devices such as data analysis, algorithmic recommendation, behavioral modeling, and digital profiling, it finely regulates individual cognition and behavior, thereby transforming the behavioral trajectory of the educated subject into a quantifiable, predictable, and controllable data stream. This causes the independence carried by value subjectivity to encounter an unprecedented technological deconstruction, with the individual’s capacity for independent value reflection and choice being deeply intervened in or even controlled by technological logic.

价值观教育是维护教育对象主体性地位和实现主流意识形态认同的重要途径，并通过基于主体间性的交互机制在主体价值认知、判断和选择中形成价值共识，从而实现价值理性的培育目标。在哲学本体论层面，价值主体性将客体价值与主体存在联结起来，价值形态的历时性演进始终围绕主体的历史性存在展开，从而在教育实践维度构建出“主体觉醒—价值认同—实践自觉”的转化机制，使个体价值选择与社会核心价值导向保持高度一致。然而，在生成式人工智能深度介入思想政治教育的情境下，数据化规训通过重塑教育对象的价值结构，逐渐形成一种新型权力运作模式，并利用数据分析、算法推荐、行为建模、数字画像等技术装置，对个体认知与行为进行精细调控，从而将教育对象的行为轨迹转化为可量化、可预测、可控制的数据流。这使得价值主体性所承载的独立性遭遇前所未有的技术性解构，个体独立进行价值反思与选择的能力被技术逻辑深度干预乃至控制。

First, the data-driven locking of the value cognition framework. In teaching practice, when information推送, knowledge acquisition, and autonomous learning are all governed by algorithms, the breadth of an individual’s value cognition is confined by the “information cocoon,” the underlying logic of cognition is reshaped by data, the subject’s critical thinking is dissolved by instant feedback mechanisms, and the freedom of value choice is co-opted by data. Second, the temporal disappearance of value reflection. The real-time data correction function of AI systems gives rise to “microparticle existence,” where individuals lose the temporal depth necessary for value sedimentation amid continuous behavioral adjustments. Internal value standards are ceded to external data indicators, leading to a humanistic dilemma of value dilution in ideological and political education. At this point, educational platforms use matching algorithms to define emotional value, teaching evaluation systems quantify morality procedurally, and the intersubjective construction of value is alienated into a process of debugging technical parameters. Third, values are shaped by data-driven algorithms. Data-driven discipline, through technical means, transforms human behavior, emotions, and thoughts into calculable, predictable, and intervenable data objects. Affective computing, through multimodal recognition mechanisms (such as facial expressions, voice, and respiration), converts complex value judgments into quantifiable data symbols, and conducts perception and analysis through psychological, behavioral, and physiological measurement methods. For example, a system dynamically assesses students’ degree of identification with specific political topics through facial expression intensity coefficients and adjusts teaching strategies accordingly. This is essentially a hypothesis of a strong correlation between biological signals and value judgments. The deconstruction of value subjectivity by data-driven discipline reveals the technological face of the crisis of modernity. However, this deconstructive effect is not the inevitable fate of technological development, but rather a consequence of humanity ceding the power of value judgment to technological systems.

一是价值认知框架的数据化锁定。在教学实践中，当信息推送、知识获取、自主学习均受算法支配时，个体的价值认知广度就会被“信息茧房”限定、认知的底层逻辑被数据化重塑，主体批判性思维被即时反馈机制消解，价值选择自由也被数据所裹挟。二是价值反思的时间性消逝。人工智能系统的实时数据矫正功能催生“微粒化生存”，个体在持续的行为调适中失去价值沉淀所需的时间纵深，内在价值尺度让渡于外部数据指标，导致思想政治教育面临价值稀释的人文困境。此时，教育平台使用匹配算法定义情感价值，教学评价系统将道德进行程序式量化，价值的主体间性建构被异化为技术参数的调试过程。三是价值观被数据化的算法形塑。数据化规训通过技术手段将人的行为、情感和思想转化为可计算、可预测、可干预的数据对象，情感计算通过多模态识别机制（如表情、声音和呼吸）将复杂价值判断转化为可量化的数据符号，并以心理、行为和生理等测量方式进行感知和分析。例如，系统通过面部表情强度系数动态评估学生对特定政治议题的认同度并据此调整教学策略，实质上是生物信号与价值判断的强相关性假设。数据化规训对价值主体性的解构揭示了现代性危机的技术面相，但解构效应并不是技术发展的必然宿命，而是人类将价值裁判权让渡给技术系统的后果。

(II) Algorithmic Cognitive Substitution Leads to the Technological Deprivation of Thinking Autonomy

（二）算法认知替代导致思维自主性的技术性剥夺

In the field of ideological and political education, maintaining thinking autonomy and independence constitutes an inherent requirement for the cultivation of modern personality and is also the cornerstone for realizing the free and comprehensive development of the human being. From the perspective of Marxist technological critique, the essence of algorithmic cognitive substitution is the cognitive extension of labor alienation in the intelligent era, thereby forming a new type of cognitive dilemma based on digital alienation. When generative AI deeply erodes the essence of ideological and political education with an alienated technological logic, the fact that instrumental rationality oversteps value rationality causes the autonomy of the subject’s thinking to be technologically deprived. The subject passively converges toward the cognitive paradigm manipulated by algorithms, thereby forming a dominating effect of technology over subjectivity — algorithmic cognitive substitution leads to the technological deprivation of thinking autonomy.

在思想政治教育场域中，保持思维自主性与独立性构成现代人格培育的内在要求，也是实现人的自由全面发展的基石。在马克思主义技术批判视域下，算法认知替代的本质是智能时代劳动异化的认知论延伸，由此形成基于数字异化的新型认知论困境。当生成式人工智能以异化的技术逻辑深度侵蚀思想政治教育本质时，工具理性对价值理性的僭越性事实使得主体思维的自主性被技术性剥夺，并被动地向算法操纵的认知范式趋同，由此技术对主体性形成宰制效应——算法认知替代导致思维自主性的技术性剥夺。

First, the technological restructuring of the cognitive architecture. On the one hand, the digital interface implements cognitive capture through the attention economy mechanism, and the algorithmic recommendation system, via the closed-loop domestication mechanism of “information cocoon — cognitive inertia — algorithmic dependence,” causes educational subjects to gradually lose their autonomous information-filtering ability, thereby forming a disciplinary effect dominated by interface cognition and a cognitive domestication effect dominated by algorithmic authority. On the other hand, the deep intervention of intelligent technology in the human thought process leads to a structural alienation between the cognitive subject and its own thinking ability, the essence of which is an extension of Marx’s theory of “alienation of labor” into the cognitive dimension in the digital age. When artificial intelligence undertakes high-level cognitive labor such as knowledge retrieval, data analysis, and value judgment, the technical tool is alienated from a cognitive aid into a dominating force, easily causing an outsourcing phenomenon of the thought process. At this point, core cognitive links such as knowledge construction, contradiction analysis, and systematic reflection are forced to be ceded to the intelligent system, resulting in a systematic degradation of the subject’s metacognitive ability; the algorithmic recommendation system, through its closed-loop domestication mechanism, reduces the subject’s autonomous cognition to a vassal of the algorithmic cognitive system, causing the algorithmization of the cognitive model. Second, the algorithmic colonization of the thought process. The algorithmic substitution of critical thinking under the dominance of intelligent technology drives the transformation of human thought from dialectical logical deduction to pattern-matching response, from value-reflective cycles to instant-feedback dependence, and from historically contextualized understanding to fragmented information processing. When educational subjects rely on intelligent devices to complete high-level thinking tasks, their metacognitive cycle falls into the trap of cognitive outsourcing, unable to extricate itself, because it is replaced by technological agency. The instant answer feedback provided by the algorithm weakens the subject’s motivation for autonomous thinking, thus easily presenting a mindset of “the algorithm is truth” and “either-or” in problem analysis. The implicit disciplinary effect of algorithmic logic on human thought can simplify complex human thinking processes into calculable and quantifiable standardized pathways, gradually weakening the autonomous thinking ability of people within the information cocoon, thereby causing the subject’s cognitive autonomy to quietly collapse under algorithmic domestication. The reasoning model of generative artificial intelligence is essentially a procedural simulation of human nonlinear thinking; its technical mimetic characteristics obscure the Marxist law of the negation of the negation, prompting the subject’s ability for contradiction analysis to tend toward flatness.

一是认知架构的技术化重构。一方面，数字界面凭借注意力经济机制实施认知捕获，算法推荐系统通过“信息茧房—认知惯性—算法依赖”的闭环驯化机制，使教育对象逐渐丧失自主信息筛选能力，进而形成界面认知主导的规训效应以及产生算法权威主导的认知驯化效应。另一方面，智能技术深度介入人类思维过程，导致认知主体与自身思维能力之间产生结构性疏离，其本质是数字时代马克思“劳动异化”理论在认知维度的延伸。当人工智能承担知识检索、数据分析、价值判断等高阶认知劳动时，技术工具从认知辅助手段异化为支配性力量，极易造成思维过程的外包化现象。此时，知识建构、矛盾分析、系统反思等核心认知环节被迫让渡给智能系统，致使主体元认知能力的系统性退化；算法推荐系统通过闭环驯化机制，使主体的自主认知沦为算法认知体系的附庸，造成认知模式的算法化。二是思维过程的算法化殖民。智能技术主导下的批判性思维的算法替代，促使人的思维从辩证逻辑推演向模式匹配响应转变，从价值反思循环向即时反馈依赖转变，从历史情境化理解向碎片化信息处理转变。当教育对象依赖智能设备完成高阶思维任务时，其元认知循环因被技术代理取代而陷入认知外包陷阱之中不能自拔，算法提供的即时答案反馈会削弱主体自主性思考的动力，从而在问题辨析中容易呈现“算法即真理”和“非此即彼”的思维定势。算法逻辑对人类思维产生的隐性规训效应，能够将复杂的人类思考过程简化为可计算和可量化的标准化路径，使人在信息茧房中逐渐弱化自主思考能力，从而导致主体认知自主性在算法驯化中悄然瓦解。生成式人工智能的推理模式本质上是对人类非线性思维的过程性模拟，其技术性的拟态特质遮蔽了马克思主义否定之否定规律，促使主体的矛盾分析能力趋向平面化。

(3) Technological mediation causes a structural transformation of intersubjectivity

（三）技术中介化造成主体间性的结构性转换

Handling the relationship between educational subject and object is an important aspect of ideological and political work, and the communication paradigm represented by intersubjectivity is a key bond for realizing value transmission and emotional connection. Generally speaking, subjectivity and interactivity are the prerequisites for intersubjectivity; “intersubjectivity regards education as communication between subjects, thereby confirming that education is the authentic mode of human existence.” In the traditional context, technological artifacts only serve as objects subject to human transformation, the subject-object relationship between the two is relatively clear, and there is no intersubjectivity based on subject interaction between humans and things. However, in the process of interaction between humans and artificial intelligence, the quasi-subject characteristics of intelligent systems enable them to gradually acquire the practical ability to autonomously transform the objective world, thereby transforming the traditional “subject-object” relationship between humans and technology into a “subject-subject (quasi-subject)” relationship, and intersubjectivity thus undergoes a structural transformation. In the scenario of ideological and political education driven by artificial intelligence, the original binary subject structure of “subject (teacher) — subject (student)” is transformed into a ternary subject structure of “subject (teacher) — subject (quasi-subject) — subject (student).” The change in interacting subjects also causes a structural transformation of intersubjectivity. First, the algorithmic reconstruction of educational relations leads to the alienation of intersubjectivity. The two-dimensional embodied interaction of “teacher-student” in traditional ideological and political education is alienated into a three-dimensional technological mediation model of “human-machine-human,” with interaction reduced to a symbolic exchange game centered on data flow. The crucial multimodal embodied interaction in interpersonal communication (such as emotional transmission through mood swings and tone changes) is weakened on the virtual interface, creating a dilemma of “interaction impoverishment.” Second, the dual migration crisis of the subjectivity structure. The quasi-subjectivization of intelligent systems gives rise to a zero-sum game in the distribution of subjectivity, causing the professional authority of the educator to migrate to the algorithmic system (e.g., the transfer of teaching evaluation authority to learning analysis systems), and the value construction of the educated to migrate to data models (e.g., the simplification of ideological dynamics into sentiment analysis curves), resulting in the phenomena of “hollowing out” and “fragmentation” of subjectivity.

处理好教育主客体关系是思想政治工作的重要方面，而以主体间性为代表的交往范式则是实现价值传递和情感联结的关键纽带。一般来讲，主体性和交互性是主体间性的前提条件，“主体间性把教育看作是主体间的交流，从而确证了教育是人本真的生存方式”。传统语境下，技术人工物仅作为客体接受人类的改造，二者之间的主客体关系较为清晰，人与物之间尚不存在基于主体交互的主体间性。然而，在人与人工智能的交互过程中，智能系统所具有的类主体性特质使其逐渐具备自主改造客观世界的实践能力，进而使传统人技之间的“主—客”关系转变“主—主（类主体）”关系，主体间性由此发生结构性转换。人工智能驱动思想政治教育场景下，原本“主（师）—主（生）”的二元主体结构转变为“主（师）—主（类主体）—主（生）”三元主体结构，交互主体的变化同样会使主体间性发生结构性转换。一是教育关系的算法化重构引发主体间性异化。传统思想政治教育中“师—生”的双向度具身交互，被异化为“人—机—人”的三向度技术中介模式，交互沦为以数据流为核心的符号交换游戏，人际交往中至关重要的多模态具身交互（如情绪波动和语调变化等情感传递）在虚拟界面被削弱而产生“交互贫困化”困境。二是主体性结构的双重迁移危机。智能系统的准主体化发展催生主体性分配的零和博弈，致使教育者的专业权威向算法系统迁移（如教学评估权让渡给学习分析系统），受教育者的价值建构向数据模型迁移（如思想动态被简化为情感分析曲线），致使主体性呈现“空心化”与“碎片化”现象。

In educational scenarios involving intelligent technology intervention, the structural transformation of intersubjectivity roughly takes two forms. On the one hand, under the effect of weak artificial intelligence, technological mediation expands the space of human subjectivity and enables it to freely transform between the virtual and the real across time and space. Canadian scholar Marshall McLuhan pointed out that technology extends the human cognitive process, “just as our senses and nervous system are extended through various media.” Technology, as a mediating existence, not only reconstructs the interactive interface between physical space and digital space but also dissolves the essential distinction between organisms and artifacts, expanding the extension of human subjectivity. On the other hand, the ontological mediating effect of strong artificial intelligence drives intelligent systems to break through the object boundary of traditional technological artifacts, thereby generating cognitive entities with quasi-subjectivity. This paradigm shift not only triggers the cognitive transfer of human subjectivity to algorithmic systems (e.g., decision-making transfer) but also leads to a fundamental change in the subjectivity structure. Through the continuous evolution of deep learning and autonomous decision-making capabilities, strong artificial intelligence gradually breaks through the object positioning of traditional technological artifacts, giving rise to a dual-structure entity that possesses both instrumental objectivity and cognitive subjectivity, forming a composite interactive system that breaks the traditional subject-object binary framework. Although it can be regarded as a “high-level object cloaked in subjectivity” or “pseudo-intersubjectivity,” “strong artificial intelligence with autonomous consciousness is no longer a tool executing programs; its will and actions no longer represent the consciousness and actions of the master or humanity.”

在智能技术介入教育场景中，主体间性的结构性转换大致存在两种形式。一方面，在弱人工智能作用下，技术中介化拓展了人的主体性空间，并能够使其在跨时空的虚实之间自由转换。加拿大学者麦克卢汉指出，技术对人类认知过程的延伸作用，“正像我们的感觉器官和神经系统凭借各种媒介而得以延伸一样”。技术作为媒介性存在，不仅重构了物理空间与数字空间的交互界面，也消解了生物体与人工物之间的本质区隔，拓展了人类主体性的外延。另一方面，强人工智能的本体论中介效应，驱动智能系统突破传统技术人工物的客体边界，进而生成具有准主体性的认知实体。这种范式跃迁不仅引发人类主体性向算法系统的认知让渡（如决策让渡），而且导致主体性结构发生根本性改变。强人工智能通过深度学习与自主决策能力的不断进化，逐渐突破传统技术人工物的客体定位，催生出一种兼具工具客体性与认知主体性的双结构存在体，形成一种打破传统主客二元框架的复合交互系统。尽管可将其视为“披着主体性外衣的高阶客体”“伪主体间性”，但“具有自主意识的强人工智能不再是一个执行程序的工具，它的意志和行动也不再是代表着主人或人类的意识与行动。”

II. Subjectivity Ethical Risks Arising from Generative Artificial Intelligence Empowering Ideological and Political Education

二、生成式人工智能赋能思想政治教育带来的主体性伦理风险

The dual nature of technology offers an important insight for the technologization of education: in practice, we must both fully exploit the positive effects of technology in promoting educational development and remain constantly vigilant against the various risks and challenges it brings. In reality, the alienation of educational technologization easily leads to a subjectivity crisis of the “humanization” of machines and the “mechanization” of humans, producing problems such as symbolic identity, alienation of interpersonal relationships, and obstacles to consensus building. Moreover, in the context of generative artificial intelligence empowering ideological and political education, the excessive promotion of one-dimensional technological instrumental rationality causes the subjectivity of educational subjects to undergo reactive alienation under the domination of technological logic, thereby generating multiple ethical risks in a state of imbalance between technology and ethics.

技术的两面性特征给教育技术化带来的重要启示在于，我们在实践中既要充分发掘技术推动教育发展的正向效能，也要时刻警醒其所带来的各种风险和挑战。现实中，教育技术化的异化极易造成机器的“人化”和人的“机器化”的主体性危机，产生符号化身份、人际关系疏离、共识凝聚阻碍等问题。不仅如此，在生成式人工智能赋能思想政治教育情境下，技术的单向度工具理性的过度张扬会使教育对象的主体性受到技术逻辑宰制而发生应激性异化，从而在技术与伦理关系失调状态下衍生多重伦理风险。

(1) Weakening of subject belief leads to disorder of value faith

（一）主体信念弱化引发价值信仰失序

Education in ideals and convictions is an important pathway for guiding students to establish a correct worldview, outlook on life, and value system. Conviction refers primarily to the subject’s deep-seated belief in a certain thought, concept, or culture; its formation and change are deeply influenced by the state of ideology. Value faith refers mainly to the value ideal or commitment that subjects collectively choose based on shared expectations of value goals; its formation depends primarily on specific ideals and convictions, or, put differently, conviction is the foundation of faith, and faith is the higher-order expression of conviction. As a necessary existence of class rule, ideological and political education systematically shapes a specific system of value faith through the value-integrating function of the dominant social ideology. “The firmness of ideals and convictions comes from the firmness of thought and theory.” The thought and theory in ideological and political education possess a strong ideological character and are an important source for systematically shaping the subject’s foundation of ideals, convictions, and values. The powerful knowledge-production capacity of generative artificial intelligence can create favorable conditions for the study of thought and theory, but the content it generates possesses a high degree of autonomy, creativity, and uncertainty. When the algorithmic model deviates from or diverges from the mainstream ideology, the intelligently generated content may transmit heterogeneous values to the educatee in the form of “soft indoctrination,” thereby shaking the foundation of their ideals and convictions and weakening their faith in the mainstream social values.

理想信念教育是引导学生树立正确世界观、人生观、价值观的重要途径。信念主要指主体对某种思想、观念、文化的深信不疑，其形成和变化深受意识形态状况的影响。价值信仰主要是指主体基于共同的价值目标期待所共同选择的价值理想或承诺，其形成主要依赖于特定的理想信念，或者说信念是信仰的基础、信仰是信念的高阶表现形式。思想政治教育作为阶级统治的必然性存在，通过社会主流意识形态的价值整合功能，系统塑造特定的价值信仰体系。“理想信念的坚定，来自思想理论的坚定。”思想政治教育中的思想理论具有浓厚的意识形态属性，是系统形塑主体理想信念和价值根基的重要源头。生成式人工智能的强大知识生产能力能够为思想理论学习创造良好条件，但其生成的内容具有高度的自主性、创造性及不确定性，当算法模型与主流意识形态产生偏差或者背离时，智能生成的内容就有可能会把异质性价值观以“软性灌输”的方式传递给受教育者，进而动摇其理想信念根基并使其弱化对社会主流价值观的信仰。

The ideological risks generated by generative artificial intelligence are inseparable from the sources of training data and technical deficiencies. On the one hand, ideological risks arise from data information carrying negative value forms. In reality, algorithmic design practice is deeply influenced by value stances such as consequentialism, virtue ethics, and deontology; an artificial intelligence entity loaded with data containing diverse ethical values, cultural forms, and subjective biases can “control social power through a technological logic that appears to control nature, shaping people’s value concepts and behavioral patterns.” [5]5 The original Chinese text cites a source here; the quotation is translated directly. For example, the training corpus of the ChatGPT large language model is primarily drawn from the Western value system, and the value orientation carried by its generated content is mainly based on mainstream Western values. If the training data used by artificial intelligence contains content that violates core socialist values, such as political stance bias, racial bias, historical nihilism, money worship, or extreme liberalism, then the content generated or recommended by the algorithm is highly likely to match such content, and the artificial intelligence entity will become an intelligent “source” of a large number of ideological risks. Once such harmful content spreads widely in the practice of ideological and political education and deeply influences the educatee’s worldview, outlook on life, and value system, the mainstream social ideology and values risk being marginalized, potentially causing disorder in value faith among the student population and weakening their positive ideals and convictions. On the other hand, ideological risks arise from technical deficiencies. First, the “filter bubble” constructed by algorithmic recommendation systems through user profiling causes the information contact surface to exhibit hypergeometric contraction and information homogenization, leading to a decline in cognitive diversity and a blunting of value-critical capacity. Second, the technical defect of AI hallucination induces “technologically unconscious” lying and fabrication of facts, creating an ontological trust crisis and undermining the rational basis of value judgment, posing a latent threat to values education. Third, cultural biases in the training data of large models may be covertly infiltrated through intelligent teaching systems and transmitted to students, causing values to degenerate from an intersubjective meaning network into an algorithmically operable parameter space, thereby laying hidden dangers for value distortion.

生成式人工智能所产生的意识形态风险，与训练数据来源和技术缺陷密不可分。一方面，源自负面价值形态的数据信息引发的意识形态风险。现实中，算法设计实践会受到功利论、德性论、义务论等价值立场的深度影响，负载着多元伦理价值、文化形态、主观偏见等数据的人工智能体，能“以一种看似控制自然的技术逻辑控制着社会权力，塑造人们的价值观念和行为模式”。例如，ChatGPT大语言模型的训练语料主要来自西方价值体系，其生成内容所负载的价值取向则以西方主流价值观为主。如果人工智能采用的训练数据中存在政治立场偏见、种族偏见、历史虚无主义、拜金主义、极端自由主义等违反社会主义核心价值观的内容，那么智能生成或算法推荐的内容也极有可能与之相匹配，人工智能体便会成为大量意识形态风险的智能“发源地”。一旦这种有害内容在思想政治教育实践中广泛扩散并深度影响受教育者的世界观、人生观、价值观，那么社会主流意识形态和价值观将有被边缘化的风险，由此在学生群体中就有可能造成价值信仰失序并弱化其正向理想信念。另一方面，技术缺陷引发的意识形态风险。其一，算法推荐系统通过用户画像构建的“过滤气泡”，促使信息接触面呈现超几何收缩和信息同质化现象，导致认知多样性衰减与价值批判力钝化。其二，AI幻觉的技术性缺陷引致“技术无意识”性的说谎和虚构事实，制造本体信任危机和动摇价值判断的理性基础，对价值观教育构成潜在威胁。其三，大模型训练数据中的文化偏见可能通过智能教学系统进行隐性渗透，并将相关的隐性偏见传导给学生，致使价值观从主体间性的意义网络蜕变为算法可操作的参数空间，从而为价值扭曲埋下隐患。

(II) Subject Thinking Inertia Leads to the Obscuring of the Spirit of Innovation

（二）主体思维惰化导致创新精神遮蔽

The cultivation of innovation capacity is not only an essential requirement of educational modernization but also a key focus for ideological and political education to fulfill its mission of forging souls and educating people. Ideological and political education takes the reconstruction of people’s ideological concepts as its value orientation, and enhancing the subject’s innovation capacity fundamentally depends on cultivating a spirit of innovation centered on innovative thinking. The spirit of innovation, innovative personality, and innovative ability are important contents of ideological and political education; ideological and political workers should play the role of soul engineers who inspire students’ creative thinking, creative imagination, and cultivate creative talent. For this reason, the importance of ideological and political education in cultivating students’ creative thinking is concentrated in aspects such as guidance by Marxist philosophy, establishing correct value orientation, shaping creative personality, and providing scientific thinking methods. However, students’ excessive reliance on intelligent technology in the process of innovative practice causes innovation elements such as problem discovery, the process of critical reflection, and original expression to be replaced by algorithmic recommendation, data modeling, and automatic generation, leading to the degeneration of innovative activity from subjective creation into technical operation. Innovative practice thus forms a pair of tension-filled contradictions between “technological enhancement” and “subjective guardianship.” From the perspective of knowledge innovation, current generative artificial intelligence is still unable to complete high-level innovative tasks; it can only simulate human thinking to engage in low-level creative work. The innovation capacity it possesses belongs to the type of “creating something from something,” and it cannot yet complete the production of new knowledge “from nothing” as humans can. Furthermore, although generative artificial intelligence can complete combinatorial innovation based on data correlation, it is mainly “low-level innovation” or “pseudo-innovation”; it still has essential limitations in breakthrough or original innovation and cannot yet meet the needs of high-level innovation.

创新能力的培育不仅是教育现代化的本质要求，更是思想政治教育实现铸魂育人使命的重要着力点。思想政治教育以人的思想观念重构为价值旨归，而提升主体的创新能力从根本上依赖于以创新思维为核心的创新精神的培养。创新精神、创新个性、创新才能是思想政治教育的重要内容，思想政治工作者应当扮演启发学生创造性思维、创造性想象力和培养创造性人才的灵魂工程师角色。正因如此，思想政治教育培养学生创造性思维的重要性，集中体现在马克思主义哲学指导、树立正确的价值导向、塑造创造性人格、提供科学思维方法等方面。然而，学生在创新实践过程中对智能技术的过度依赖，使得问题发现、思辨过程、原创表达等创新要素被算法推荐、数据建模、自动生成的技术取代，导致创新活动从主体性创造退化为技术性操作，创新实践在“技术增强”与“主体性守卫”之间形成一对张力性矛盾。从知识创新角度来讲，目前生成式人工智能尚难以完成高阶创新任务，只能模拟人类思维从事低阶创造工作，其所具备的创新能力属于“有中生有”的类型，还无法像人类那样完成“从无到有”的新知识生产。进一步讲，生成式人工智能虽能完成基于数据关联的组合式创新，但主要以“低阶创新”或“伪创新”为主，其在突破性或原创性创新层面仍存在本质局限，尚不能满足高阶创新需要。

Faced with the convenience and high efficiency brought by the deep integration of intelligent technology and ideological and political education, if the educatee’s discernment is insufficient or self-control is lacking, they are prone to excessive dependence on the artificial intelligence entity, thereby obscuring and weakening the positive role of Marxist scientific thinking and creative personality, inducing and amplifying the subject’s thinking inertia, leading to the proliferation of superficial thinking and weak thinking, and thus giving rise to a widespread crisis of the obscuring of the spirit of innovation. On the one hand, excessive technological dependence causes a reduction in the subject’s cognitive level and a structural atrophy of critical thinking. Unconditional reliance on generative artificial intelligence for knowledge production easily fosters a trend of unthinking intellectual “take-ism,” gradually weakening the subject’s capacity for reflection, judgment, and creativity, ultimately leading to the inertia of the subject’s thinking and the diminishment of the critical spirit. Algorithmic path dependence causes alienation in the subject’s cognitive process; the instant provision of answers by artificial intelligence enables the subject to bypass the necessary cognitive stage of “problem suspension – dialectical reflection,” causing their thinking inertia to become increasingly entrenched. Continuous reliance on artificial intelligence to complete knowledge integration and value judgment leads to the gradual degradation of self-monitoring and self-regulation mechanisms, resulting in the continuous atrophy of the subject’s metacognitive ability. The machine-dominated mode of knowledge production is essentially an algorithmic reshaping of the dialectical cognitive process, tempting the user to degenerate from an active practical subject into a passive consumer of technological products, resulting in the objectification of the subject. On the other hand, the proliferation of “intelligent creation” deeply erodes the subject’s innovative initiative. The subject’s innovative initiative is replaced by the proliferation of “intelligent creation”; the innovation cycle of “bad money driving out good” weaves a low-level “innovation cocoon,” leading to a lack of creative motivation, a weakening of creative ability, and a dilution of innovation awareness. Thus, the reinforcement of low-level innovation coexists with the crisis of innovation homogenization; innovation degenerates into a technical game of parameter adjustment, and the innovation level is locked within a limited space, resulting in a creative poverty of knowledge production.

面对智能技术与思想政治教育深度融合带来的便利性和高效率，如若教育对象的辨识力不够或自控力不足，则容易对人工智能体产生过度依赖，从而遮蔽和弱化马克思主义科学思维和创造性人格的积极作用，催生和放大主体的思维惰性并导致浅层思考、弱思考泛滥，由此产生普遍性的创新精神遮蔽危机。一方面，过度性技术依赖引发主体认知降维，以及批判性思维出现结构性萎缩。无底线地依赖生成式人工智能进行知识生产，易催生不加思索的知识“拿来主义”之风，使主体的思考力、判断力与创造力渐趋弱化，最终导致主体思维惰化、批判精神消减。算法路径依赖会使主体认知过程出现异化，人工智能的即时答案供给会使主体跨越“问题悬置—辩证反思”的必要性认知阶段，导致其思维惰性不断固化；持续依赖人工智能完成知识整合与价值判断，使自我监控与自我调节机制逐渐退化，造成主体元认知能力不断萎缩。机器主导的知识生产模式本质上是对辩证认知过程的算法化重塑，诱使使用者从能动的实践主体沦为技术产品的被动消费者而产生主体客体化。另一方面，泛滥的“智能创造”对主体创新能动性的深度侵蚀。主体的创新能动性被泛滥的“智能创造”所取代，“劣币驱逐良币”的创新循环编织出低阶性“创新茧房”，进而导致主体创造动力匮乏、创造能力弱化和创新意识淡化。由此，低阶创新不断强化和创新同质化危机并存，创新蜕变为参数调整的技术游戏，创新能级被锁定在有限空间内，造成知识生产的创造性贫困。

(III) Subject Perception Virtualization Leads to the Dilution of Humanistic Emotion

（三）主体感知虚化造成人文情感稀释

Humanistic emotion is essentially an embodied representation of intersubjectivity, a vital bond for sustaining social ethical relations and realizing individual development. The subject’s perceptual capacity is the epistemological prerequisite for generating emotional capacity, playing an irreplaceable role in the transmission of emotion in ideological and political education. Ideological and political education is an educational endeavor imbued with thoughtfulness and humanism; it places particular emphasis on emotional exchange and humanistic care derived from firsthand experience based on actual perception, thereby continuously promoting the emotional development and moral cultivation of the educational subject. The embodied emotional cultivation paradigm of traditional ideological and political education relies on “intersubjective dialogue” in physical presence. Through situated perception and concrete interaction, it shapes a distinct advantage of presence in which subject and object merge, enabling a rational transition from sensory stimulation and emotional arousal to value reflection and behavioral internalization. This subject-object fusion makes perception both authentic and multidimensional, facilitating the expression and transmission of humanistic emotion between teachers and students, as well as the realization of emotionally warm educational exchange. However, the artificial intelligence entity itself does not possess the capacity for emotional experience; it is merely a rational interactive subject. The AI entity’s cognition and perception of emotion are built upon rational objective data, and its emotional simulation is essentially an algorithmic reproduction of the statistical patterns of emotion’s external manifestations. It lacks the genuine emotional capacities that humans possess, such as joy, anger, sorrow, pleasure, empathy, feeling, and aesthetic sensibility. On the one hand, the virtual teacher generated by AI systems, through data analysis and affective computing, produces “standardized empathy,” which easily severs the organic connection between emotional expression and intrinsic value, inducing a crisis of symbolization in emotional exchange amid the illusions of “simulacra replacing truth” and “dislocation between simulacra and truth.” On the other hand, although current AI technology can simulate sensory stimulation to a certain degree, it cannot reproduce embodied elements such as the conduction of temperature and the arousal of emotion in actual perception, thereby leading to the dissolution of embodied cognition.

人文情感本质上是主体间性的具身化表征，是维系社会伦理关系和实现个体发展的重要纽带，其中主体感知能力则是生成情感能力的认识论前提，在思想政治教育中发挥着不可替代的情感传递作用。思想政治教育是一项充满思想性和人文性的教育工作，尤为注重基于实际感知的亲身体验带来的情感交流和人文关怀，以此不断促进教育对象的情感发展和道德培养。传统思想政治教育的具身化情感培育范式依托物理在场的“主体间对话”，能够以情境化感知和具象化互动塑造出明显的主客交融的在场性优势，实现从感官刺激和情绪唤醒到价值反思和行为内化的合理转换。这种主客交融让感知既真实又立体，有利于师生之间人文情感的表达和传递，以及实现具有教育温度的情感交流。然而，人工智能体本身并不具备情感体验能力，仅是一种基于理性的交互主体。人工智能体对情感的认知与感知是建立在理性的客观数据之上，其情感模拟本质是算法对情感外显行为在统计学意义上的复现过程，没有喜怒哀乐、共情、感情、美感等人类所具有的真实情感能力。一方面，人工智能系统生成的虚拟教师通过数据分析和情感计算生成的“标准化共情”，容易割裂情感表达与内在价值的有机联系，在“拟像替代真相”和“拟像与真相错位”的幻象中诱发情感交流的符号化危机。另一方面，现有人工智能技术虽然在一定程度上能够模拟感官刺激，但无法复现实际感知的温度传导和情绪唤起等具身性要素，从而导致具身认知的消解。

The virtual teacher lacks genuine emotional feedback and the capacity for emotional empathy. Excessive reliance on it will seriously impede the embodied cultivation of students’ humanistic emotions. The proliferation of virtual scenes easily leads to phenomena such as emotional cognitive disconnection and the “decorporalization” of empathic experience, thereby resulting in “emotional blunting” when facing real social problems. To a certain extent, this can dissolve the human-centered nature of ideological and political education. The rich humanistic emotional exchange and personality education that originally existed between teachers and students may be replaced by cold, rigid, and insensate symbols. The meaning and value of the subject’s existence are gradually dissolved and assimilated by the machine, spreading toward the extreme direction of the virtualization of subjective perception. Although the human-like capabilities of AI technology are continuously strengthening, it lacks the unique humanistic elements of free will, inspiration, emotion, and ethics, which limits its positive role in emotional understanding, spiritual empathy, and mental feeling. Human-computer interaction merely simulates human behavior through forms of judgment and reasoning such as data, algorithms, code, and models; it can ultimately never understand the human spiritual world or the essence of emotion. The excessive manipulation of emotional education by AI causes “the value and meaning that should be possessed to be filtered out, the collision of thought and values to be simplified into a procedural, homogenized algorithmic operation, and the originally wisdom-filled, warm interpersonal relationships to become a cold, mechanical process of exchanging information symbols. This inevitably leads to a decline in individual empathic ability and a reduction in the sense of experience.” AI cannot replace human emotional power, imagination, or moral power. The alienation of human-computer relations leads to the homogenization of humans and machines or a situation in which humans become dependent on machines, turning ideological and political education, which should be full of “human touch,” into a cold and impersonal “machine education.”

虚拟教师缺乏真实情感反馈和情感共情能力，对其过度依赖会严重阻滞对学生的人文情感的具身化培养，泛滥的虚拟场景极易导致情感认知断层、共情体验“去身体化”等现象，由此在面对真实社会问题时容易出现“情感钝化”。这在一定程度上会消解思想政治教育的人本性，原本师生之间所具有的丰富人文情感交流和人格教育有可能会被冷漠、呆滞、无感的符号所代替，主体的存在意义和价值逐渐被机器消解和同化，并朝着主体感知虚化的极端方向蔓延。尽管人工智能技术的类人能力不断增强，但缺乏自由意志、灵感、情感、伦理等人所独有的人文要素，使其在情感理解、心灵共情、精神感受等方面的积极作用依旧有限。人机交互仅仅通过数据、算法、代码、模型等判断推理形式来模拟人类行为，终究无法理解人类的精神世界和情感本质。人工智能对情感教育的过度操控，使得“本该拥有的价值与意义被过滤，思维、价值观的碰撞被简化为程序化、单一化的算法运行，原本充满智慧、富有温度的人际关系变为冰冷、机械的信息符号交互过程，这必然会导致个体共情能力下降以及体验感的降低。”人工智能无法代替人类的情感力、想象力、道德力，人机关系异化导致人机同质化或人依附于机器的情景，会让本应充满“人情味”的思想政治教育成为冷冰冰的“机器教育”。

(4) The Dilution of Subject Responsibility Consciousness Gives Rise to Moral Behavioral Anomie

（四）主体责任意识淡化催生道德行为失范

A sense of responsibility plays an important role in shaping individual moral behavioral norms, and guiding and cultivating the subject’s responsibility consciousness is a key aspect of ideological and political education. The instrumental rationality of intelligent technology’s transgression of subjectivity not only enhances the controllability of the educational process through means such as data and algorithms but also creates a predicament of the dissolution of subjectivity at the level of responsibility ethics, thereby forming a structural tension between technological empowerment and responsibility ethics. When technological rationality breaks through the boundary of instrumentality and evolves into a dominating force, the technological deprivation of human autonomy by intelligent technology plunges the subject into a digital virtual mirror image. The subject’s responsibility consciousness is reconstructed by algorithmic logic, degenerating into a passive response to external discipline and losing the practical foundation of “self-legislation,” thereby triggering a crisis of responsibility ethics. In the educational field empowered by generative AI, the multi-dimensional responsibility network constituted by subjects such as system developers, teachers, administrators, and students exhibits characteristics of the dispersal of responsible subjects, the blurring of responsibility boundaries, and the fragmentation of responsibility cognition, thereby creating “responsibility blind spots” or “responsibility vacuums.” For example, system developers focus on product performance and design defect responsibility, while teachers focus on classroom application and usage responsibility. In the network ecology where multiple responsible subjects coexist, the phenomena of responsibility suspension and responsibility dilution intertwine, forming a chain-reaction mechanism of responsibility diffusion (i.e., the “bystander effect”), which leads to a series of moral behavioral anomie behaviors as the subject’s responsibility consciousness is deconstructed technologically.

责任感对个人道德行为规范具有重要的形塑作用，而对主体责任意识的引导和培养是思想政治教育的重要方面。智能技术的工具理性对主体性的僭越，既通过数据、算法等手段增强了教育过程的可控性，又在责任伦理层面制造了主体性消解的困境，从而在技术赋权与责任伦理之间形成一种结构性张力。当技术理性突破工具性边界并演变为支配性力量时，智能技术对人的自主性的技术性剥夺会使主体深陷数字化的虚拟镜像之中，主体责任意识被算法逻辑重构而沦为外部规训的被动反应并丧失“自我立法”的实践根基，由此引发责任伦理危机。在生成式人工智能赋能的教育场域中，系统开发者、教师、管理者、学生等主体构成的多维责任网络，呈现出责任主体离散化、责任边界模糊化和责任认知碎片化的特征，从而产生“责任盲区”或“责任真空”，如系统开发者关注产品性能和设计缺陷责任、教师关注课堂应用和使用责任。在多元责任主体共生的网络生态中，责任悬置与责任稀释现象相互交织，形成连锁反应的责任分散机制（即“旁观者效应”），从而导致主体责任意识在被技术性解构的过程中衍生出系列道德失范行为。

First, moral misconduct arising from platform anonymization. Platforms strip users of their real identity markers through technical means, constructing a seemingly free and equal virtual space, yet imperceptibly providing educated subjects with a “digital mask” to evade accountability. Anonymous intelligent platforms may weaken moral responsibility, inducing educated subjects to evade accountability on the grounds that “virtual behavior has no real-world consequences.” For instance, the incidence of misconduct such as verbal violence by students in platform discussion areas is far higher than in physical classrooms, which to a certain extent confirms the negative effect of identity concealment or ambiguity on the suspension of subject responsibility. Second, ethical harm caused by the computationalization of morality. When moral evaluation is transformed into data indicators in an intelligent environment, it may also fall into the misconception of “computationalization of virtue,” leading students to simplify learning outcomes into “data markers.” When moral behavior is reduced to quantifiable “data markers” (such as interaction frequency and emotional fluctuations), students easily fall into the cognitive trap of “dataism,” thereby alienating moral practice into a technical operation of “data score-chasing.” For example, in pursuit of platform points or algorithmic recommendations, they deliberately create high-frequency interactions, fabricate emotional feedback, or even reduce moral reflection to an exam-taking strategy of “matching the standard answer.” Third, a crisis of academic integrity. When technological mediation breaks through the boundary of instrumentality to reconstruct the logic of knowledge production, the subjective foundation of academic creation is shaken and creative labor tends toward alienation, leading to a paradigm shift in the perception of responsibility from “author-centered” to “human-machine hybrid” and even “machine-dominated.” Currently, the widespread application of generative artificial intelligence in the academic field has spurred an efficiency revolution in scenarios such as thesis writing, design creation, and report writing, but it has also triggered academic misconduct including plagiarism, cheating, fabrication, and copyright disputes. In particular, functions such as intelligent generation, intelligent rewriting, and intelligent polishing blur the boundary of originality in academic creation, giving rise to academic misconduct such as direct copying, covert collage, and AI-collaborated fabrication. These phenomena originate from the effect of responsibility perception drift caused by the black-box operation of subject responsibility by technology, causing academic subjects to exhibit a clear tendency toward denial and indifference when facing issues such as fact-checking, copyright attribution, and data authenticity. Fourth, a crisis of subject personality degradation. In the digital education field dominated by artificial intelligence technology, the proliferative construction of subjects’ virtual identities and the continuous shrinkage of their real identities form a tension, causing digital personality to exert a strong squeezing effect on real personality. Real personality gradually weakens or loses its capacity for moral responsibility under technological discipline, becoming degraded or alienated, thereby leading to the malformation of personality development and the impoverishment of the spiritual world, running counter to the free and comprehensive development of the human person. On the one hand, digital twins cause self-personality to split between the virtual and the real; the coexistence of multiple personas of the subject leads to personality fragmentation and a rupture in self-reflection, making it difficult for the same subject online and offline to maintain a stable system of moral principles. On the other hand, in the process of digital reconstruction where subject personality “shifts from the real to the virtual,” the disorderly expansion of virtual identity leads to the hollowing out of real personality.

一是平台匿名化带来的道德失范行为。平台通过技术手段剥离用户真实身份标识，构建起看似自由平等的虚拟空间，却无形中为教育对象提供了逃避问责的“数字面具”。匿名化智能平台可能削弱道德责任感，诱使教育对象以“虚拟行为不具现实后果”为由逃避问责，如学生在平台讨论区的语言暴力等失范行为发生率远高于实体课堂，在一定程度上印证了身份隐匿或模糊对主体责任悬置所产生的负面作用。二是道德计算化导致的伦理危害。在智能环境下将道德评价转化为数据指标时，也有可能陷入“德性计算化”的误区，导致学生将学习效果简化为“数据标识”。当道德行为被简化为可量化的“数据标识”（如互动频率和情感波动等）时，学生就容易陷入“数据主义”认知陷阱，从而将道德实践异化为“数据刷分”的技术操作。例如，为追求平台积分或算法推荐，刻意制造高频互动、伪造情感反馈，甚至将道德反思简化为“符合标准答案”的应试策略。三是学术诚信危机。当技术中介突破工具性边界来重构知识生产逻辑时，学术创作的主体性根基就会发生动摇并使创作劳动趋于异化，导致责任认知从“作者中心”向“人机混合”乃至“机器主导”的范式迁移。当前，生成式人工智能在学术领域的广泛应用，催生了论文撰写、设计创作、报告写作等场景的效率革命，但同时也引发了抄袭、作弊、造假、版权争议等学术失范问题，尤其是智能生成、智能改写、智能润色等功能模糊了学术创作的原创性边界，催生直接复制、隐性拼贴、AI协同造假等学术失范行为。这些现象根源于主体责任被技术黑箱操作所造成的责任认知漂移效应，使得学术主体在面对事实核查、版权归属、数据真实性等问题时，表现出明显的否定性态度和漠视性倾向。四是主体人格矮化危机。在人工智能技术主导的数字化教育场域中，主体虚拟身份的增殖性建构与现实身份的持续性萎缩形成张力，促使数字人格对现实人格形成强烈的挤压效应，现实人格在技术规训中逐步弱化或丧失道德担当能力而被矮化或异化，由此导致人格发展的畸形化与精神世界的贫瘠化，从而与人的自由全面发展背道而驰。一方面，数字孪生致使自我人格在虚实之间出现分裂，主体的多重人设并存导致人格碎片化及自我反思的断裂，容易使线上线下的同一主体人格处于严重的割裂状态并难以形成稳定的道德原则体系。另一方面，在主体人格实现“脱实向虚”的数字化重构过程中，虚拟身份的无序扩张导致现实人格的空心化。

III. Pathways for Resolving the Ethical Risks Arising from Generative Artificial Intelligence Empowering Ideological and Political Education

三、生成式人工智能赋能思想政治教育衍生的伦理风险的化解路径

The practice of generative artificial intelligence empowering ideological and political education is a dialectical process in which positive efficacy and negative risks coexist. We should seek a scientific path of ethical governance based on the laws of ideological and political education and the characteristics of artificial intelligence technology. AI-empowered ideological and political education must highlight the subjective position of the human person; it should both promote technological development with a humanistic concept and integrate technology into educational practice with a responsible attitude. In the New Era, to prevent and resolve the ethical risks brought by generative artificial intelligence to ideological and political education, we must base ourselves on implementing the fundamental task of fostering virtue through education [7]7 “Fostering virtue through education” (立德树人) is a fundamental task of education in China, emphasizing the cultivation of moral character and the nurturing of talent., adhere to the basic principle of “people-centered, intelligence for good,” and bring the technical rationality of artificial intelligence within the framework of the laws of ideological and political education, so that technological efficacy is released in an orderly manner according to the value rationality of education.

生成式人工智能赋能思想政治教育实践，是一个正面效能与负面风险并存的辩证性过程，我们应根据思想政治教育规律和人工智能技术特性找寻科学的伦理治理之道。人工智能赋能思想政治教育必须突出人的主体性地位，既要以人本主义理念推动技术发展，也要以负责任的态度将技术融入教育实践。新时期，防范化解生成式人工智能对思想政治教育带来的伦理风险，要立足落实立德树人根本任务，坚持“以人为本、智能向善”的基本原则，将人工智能的技术理性纳入思想政治教育规律框架之下，使技术效能按照教育的价值理性进行有序释放。

(I) Consolidating the Foundation of Value Guidance in Ideological and Political Education through the Construction of Ideological Security

（一）以意识形态安全建设夯实思想政治教育的价值引领基石

The mainstream values of a society profoundly shape the ethical characteristics of its mainstream ideology, and the ethical qualities of the mainstream ideology in turn deeply influence the state of the society’s mainstream values. Ideological and political education bears the arduous task of disseminating society’s mainstream values. In practice, it must take the construction of ideological security as a lever, compelling the technical rationality of artificial intelligence to serve the value rationality of ideological and political education, “improving the capacity for online education, and solidly carrying out school ideological and political work and ideological work in the internet era.” [8]8 This is a quotation from a speech by Xi Jinping, emphasizing the importance of ideological and political work in the internet age. First, we must continuously strengthen the leading power of mainstream ideology, ensuring that the value output of artificial intelligence conforms to society’s mainstream values. Adhere to the Party’s absolute leadership over ideological work in the education system; establish and improve a corpus that aligns with China’s mainstream ideology and values, especially by deeply embedding Marxist value concepts, core socialist values, and fine traditional Chinese culture into artificial intelligence systems, and enhancing the value-sensitive design capability of algorithms; strengthen the purification of cyberspace to root out the soil for the spread of harmful ideas and erroneous trends of thought; and reinforce proactive data “feeding” and content推送 based on mainstream social values, so that generative artificial intelligence, in the process of empowering ideological and political education, becomes a booster and multiplier for consolidating the mainstream ideological front. At the same time, relevant departments may also explore building intelligent platforms specifically targeting ideological and political work, and embed them into other existing major large language models in the form of public responsibility, thereby creating a favorable platform environment for ideological and political education. Second, explore the establishment of a full-process ideological supervision mechanism to ensure that every link of AI-empowered ideological and political education meets the requirements of mainstream social values. The value characteristics of artificial intelligence technology have clear plasticity, largely enabling the practical feasibility of alignment with specific values. Therefore, in the design process of AI entities, we must adhere to the principle of value alignment based on a holistic and systematic view, effectively embedding mainstream social value content and relevant value principles into every link of artificial intelligence, including data collection and processing, machine learning, algorithm design, content generation, and many other aspects, and establish systematic and agile mechanisms for supervision, feedback, and correction before, during, and after the event. At the same time, establish an ethical review committee for intelligent education technology targeting ideological and political education, formulate a negative list system covering links such as data collection, algorithm application, and human-computer interaction, and regularly conduct ethical impact assessments and practical guidance for technological applications.

一个社会的主流价值观深刻形塑着社会主流意识形态的伦理特性，而社会主流意识形态的伦理特质反过来又深度影响着社会主流价值观的状态。思想政治教育肩负着传播社会主流价值观的艰巨任务，在实践中要以意识形态安全建设为抓手，驱使人工智能的技术理性服务于思想政治教育的价值理性，“提高网络育人能力，扎实做好互联网时代的学校思想政治工作和意识形态工作。”一是要不断增强主流意识形态的引领力，确保人工智能的价值输出符合社会主流价值观。坚持党对教育系统意识形态工作的绝对领导，建立和完善符合我国主流意识形态和价值观的语料库，特别是要把马克思主义价值理念、社会主义核心价值观以及中华优秀传统文化深度嵌入人工智能系统当中，以及提升算法的价值敏感性设计能力；加强对网络空间的正本清源，净化不良思想和错误思潮的传播土壤，强化基于社会主流价值观的主动性数据“投喂”和内容推送，使生成式人工智能在赋能思想政治教育过程中成为巩固主流意识形态阵地的助推器和倍增器。同时，相关部门也可以尝试探索建设专门针对思想政治工作的智能平台，并以公共责任的形式将其嵌入现有其他主要大语言模型之中，为思想政治教育营造良好的平台环境。二是探索建立全流程覆盖的意识形态监管机制，确保人工智能赋能思想政治教育的各个环节符合社会主流价值观的要求。人工智能技术的价值特性具有明显的可塑性，很大程度上能够实现与特定价值对齐的现实可行性。因此，在人工智能体的设计过程中，要坚持基于整体观和系统观的价值对齐原则，将社会主流价值内容和相关价值原则有效嵌入人工智能的各个环节，包括数据采集和处理、机器学习、算法设计、内容生成等诸多方面，并在事前、事中和事后建立系统敏捷的监督、反馈和纠错机制。同时，建立针对思想政治教育的智能教育技术伦理审查委员会，制定包含数据采集、算法应用、人机交互等环节的负面清单制度，定期开展技术应用的伦理影响评估和实践指导工作。

(II) Strengthening the Construction of a New Educational Mechanism Oriented toward Innovation Capability

（二）加强以创新能力为导向的新型教育机制建设

Innovation capacity has increasingly become an important criterion for measuring the value of human talent; the unique higher-order innovation capacity of human beings is a key aspect in determining the essential difference between humans and machines, while the infinite knowledge-production capacity of artificial intelligence and its limited innovation capacity form a stark contrast. In an era when artificial intelligence technology is profoundly reshaping ideological and political education, it is necessary to establish a dialectical framework for understanding technology, which must both make good use of the instrumental rationality of intelligent technology to enhance innovation efficiency and, more importantly, guard against the risk of the dissolution of agency caused by excessive reliance on technology. Therefore, traditional modes of innovation face the demand for transformation under the impact of technology, and it is essential to strengthen the construction of new types of innovation capacity based on human-machine interaction. In particular, it is necessary to adhere to an orientation guided by the spirit of innovation, to reinforce education in the Marxist view of innovation at the cognitive and practical levels, and to improve the corresponding system of innovation evaluation standards. Through a series of corrective measures, the degradation of the subject’s innovation capacity and the risks brought about by technological addiction must be rectified. First, return to the fundamental stance of Marxist critique of technology: while making good use of intelligent technology to enhance educational effectiveness, it is also necessary to safeguard cognitive autonomy through institutional design, embed necessary reflective design within the technological system, maintain the continuous reproduction of the subject’s critical thinking, and construct a human-oriented digital-intelligent educational ecology. Second, carry out education and training work based on general knowledge of artificial intelligence technology application, with a focus on enhancing the intelligent literacy of educated subjects. Artificial intelligence literacy is mainly composed of three parts: the awareness to select and use artificial intelligence, the ability to flexibly operate artificial intelligence, and the responsibility to use artificial intelligence in a standardized manner. In practice, the goal of shaping a new type of innovation spirit should be integrated throughout the “ideological and political education + artificial intelligence” education and training work, placing education in responsibility ethics in an important position, cultivating students’ comprehensive ability to use intelligent technology for innovation, especially higher-order innovation capacity assisted by intelligent tools. Third, strengthen the education assessment and evaluation mechanism with innovation capacity as the core value orientation. As an important means for ideological and political education to cultivate innovative talent, education assessment and evaluation play a key guiding function and incentive role in the educational process. The low-order innovation characteristics of artificial intelligence determine the necessity of stimulating the subject’s innovation vitality and raising the level of innovation. In practice, attention should be paid to increasing the weight of innovation in education assessment and evaluation, setting key evaluation objectives, indicators, methods, records, and feedback around innovative thinking and innovative practice, thereby achieving a virtuous cycle between innovation capacity cultivation and assessment evaluation. Fourth, construct a new type of cultural ecology based on intelligent technology assisting subject innovation, gather ethical centripetal force around innovation principles and innovation norms, and focus on cultivating the subject’s innovation awareness, innovation capacity, and innovation responsibility. Against the background of the era when intelligent technology deeply empowers education, the innovation of educational mechanisms is shifting from single knowledge transmission to the cultivation of comprehensive literacy with innovation capacity as the core. Constructing a new type of cultural ecology with innovation as the core has become a key proposition for educational modernization. In practice, the cultivation of the educated subject’s higher-order innovation capacity should be taken as the core task, accelerating the construction of an innovation culture based on human-machine collaborative innovation capacity. By reshaping innovation concepts, innovation principles, and corresponding ethical norms, a broad innovation culture atmosphere and innovation fashion should be formed within the education field and even throughout society, thereby promoting the comprehensive leap in the subject’s innovation literacy, including innovation awareness, capacity, and responsibility.

创新能力越来越成为衡量人才价值的重要标准，人类独有的高阶创新能力是判定人机本质差异的重要方面，而人工智能的无限知识生产能力和有限的创新能力形成鲜明对比。在人工智能技术深度重塑思想政治教育的时代，需要确立辩证的技术认知框架，既要善用智能技术的工具理性提升创新效率，更要警惕对技术过度依赖导致的主体性消解风险。因此，传统创新方式在技术的冲击下面临转型要求，必须加强基于人机交互的新型创新能力建设，特别是要坚持以创新精神为导向，在认知和实践层面强化马克思主义创新观教育以及完善相应的创新评价标准体系，并通过一系列矫正手段来纠偏技术成瘾带来的主体创新能力退化现象和风险。一是回归马克思主义技术批判的根本立场，既要善用智能技术增强教育效能，又要通过制度设计守护认知自主性，在技术系统中嵌入必要的反思性设计，保持主体批判性思维的持续性再生产，构建人本导向的数智化教育生态。二是开展基于通识性人工智能技术运用的教育培训工作，着力提升教育对象的智能素养。人工智能素养主要由选择运用人工智能的意识、灵活操作人工智能的能力和规范使用人工智能的责任三大部分组成。在实践中，应将塑造新型创新精神的目标贯穿于“思想政治教育+人工智能”的教育培训工作当中，把责任伦理教育放在重要位置，培养学生运用智能技术进行创新的综合能力，特别是智能工具辅助下的高阶创新能力。三是强化以创新能力为核心价值导向的教育考核评价机制。教育考核评价作为思想政治教育培养创新人才的重要手段，在教育过程中起着关键的导向功能和激励作用。人工智能的低阶创新特性决定了激发主体创新活力和提升创新层级的必要性，现实中应该注重创新在教育考核评价中的权重，围绕创新思维和创新实践设置关键性评价目标、指标、方法、记录和反馈，由此在创新能力培养与考核评价之间实现良性循环。四是构建基于智能技术辅助主体创新的新型文化生态，围绕创新原则和创新规范凝聚伦理向心力，着重培育主体的创新意识、创新能力和创新责任。在智能技术深度赋能教育的时代背景下，教育机制的创新正在从单一的知识传授转向以创新能力为核心的综合素养培育，构建以创新为核心的新型文化生态已成为教育现代化的关键命题。实践中，应以培养教育对象的高阶创新能力为核心要务，加快推进基于人机协同创新能力的创新文化建设，通过重塑创新理念、创新原则及相应的伦理规范，在教育领域乃至全社会形成广泛性的创新文化氛围和创新风尚，以此推动主体创新意识、能力与责任等创新素养的全面跃升。

(3) Improving the Humanistic Environment for Intelligent-Assisted Ideological and Political Education

（三）改善智能辅助思想政治教育的人文环境

Xi Jinping pointed out: “Ideological and political work is fundamentally work done with people; it must revolve around students, care for students, and serve students, continuously improving students’ ideological level, political awareness, moral character, and cultural literacy, enabling students to become comprehensively developed talent.” [9]9 Xi Jinping, “Ideological and Political Work Is Fundamentally Work Done with People” (December 7, 2016), in *Excerpts from Xi Jinping’s Discourses on the Overall Strengthening and Improvement of Party Building*, Central Party Literature Press, 2021, p. 283. Ideological and political education carries the core tasks of ideological and political work. In educational practice, it focuses on “the work of shaping people’s spiritual world and invigorating their spirit,” emphasizing the transformation of people’s spiritual world and soul, thereby highlighting its humanistic characteristics. In educational practice, it is necessary to adhere to the humanistic value concept of putting people first, strengthen the humanistic care and human-centered design of digitalization and intelligence, and rationally integrate the instrumental rationality of intelligent technology with the value rationality of ideological and political education, so that generative artificial intelligence, in the process of empowering ideological and political education, possesses not only precision, breadth, and effectiveness but also reflects warmth, emotion, and sentiment. First, explore effective pathways for embedding emotional intelligence based on humanistic values into intelligent technology, with a focus on enhancing the comprehensive technical capabilities of emotional computing, emotional perception, emotional recognition, emotional simulation and expression. Emphasize the comprehensive application of cross-modal and multi-media means that combine the virtual and the real, complement human and machine, immerse in scenarios, integrate emotion and reason, innovate discourse, and unify knowledge and action. Continuously improve the emotional interaction function between humans and machines based on efficient intelligent empathy, promoting the realization of the humanistic value and humanistic care of intelligent ideological and political education. According to the laws of education in ideological and political education, integrate technologies such as virtual reality (VR), augmented reality (AR), mixed reality (MR), and extended reality (XR) into intelligent design, creating educational narrative scenes that achieve panoramic reproduction and the fusion of scene and emotion. For example, by constructing an intelligent education humanistic care mechanism, effectively embed the technical design of multimodal emotion recognition and regulation into the entire process of human-machine interaction, preventing the emotional alienation caused by the subject’s excessive reliance on technology. Second, explore benign human-machine interaction models in the context of ideological and political education, ensuring the educational subject status of both educators and educatees. The empowerment of ideological and political education by generative artificial intelligence technology should be a benign human-machine collaborative education model under human leadership, rather than an education model of alienated interpersonal relationships under machine overreach. In practice, actively advocate for the return of human agency, fully leverage the main force role of teachers as humanistic emotional educators, construct a new type of benign interaction model based on the tripartite subject of “teacher-machine-student,” organically integrate the humanistic warmth of education and teachers with the work efficiency of machines, and use human wisdom and emotion to guide intelligent technology to better serve the cause of ideological and political education.

习近平指出：“思想政治工作从根本上说是做人的工作，必须围绕学生、关照学生、服务学生，不断提高学生思想水平、政治觉悟、道德品质、文化素养，让学生成为全面发展的人才。”思想政治教育承载着思想政治工作的核心任务，在教育实践中专注于“做人的精神世界塑造和精气神提振的工作”，强调对人的精神世界和灵魂的改造，从而突出人文性特点。在教育实践中，必须坚持以人为本的人文价值理念，强化数字化和智能化的人文关怀和人本设计，合理融合智能技术的工具理性和思想政治教育的价值理性，让生成式人工智能在赋能思想政治教育的过程中，既有精度、广度和效度，又能体现温度、情感和情怀。一是探索将基于人文价值的情感智能嵌入智能技术的有效路径，着力提升情感计算、情感感知、情感识别、情感模拟与表达的综合技术能力，注重虚实结合、人机互补、场景沉浸、情理融通、话语创新和知行合一的跨模态和多介质手段的综合运用，不断完善人机之间基于高效智能化共情的情感交互功能，推动实现思想政治教育智能化的人文价值和人文关怀。根据思想政治教育的育人规律，将虚拟现实（VR）、增强现实（AR）、混合现实（MR）、扩展现实（XR）等技术融入智能设计之中，打造全景再现和情景交融的教育叙事场景。例如，通过构建智能教育人文关怀机制，将多模态情感识别和调节的技术设计有效嵌入人机互动的全过程之中，防止主体对技术过度依赖导致的情感异化。二是探索思想政治教育场景下良性的人机交互模式，确保教育者和受教育者的教育主体地位。生成式人工智能技术赋能思想政治教育，应当是人类主导下的良性人机协同教育模式，而非机器越位下的人际关系异化的教育模式。在实践中，要积极倡导人的主体性的回归，充分发挥教师作为人文情感教育的主力军作用，构建基于“师—机—生”三元主体的新型良性交互模式，让教育和教师的人文温度与机器的工作效度有机融合，用人的智慧和情感引领智能技术更好地服务于思想政治教育事业。

(4) Establishing and Improving a Moral Responsibility Education Mechanism Based on Intelligent Scenarios

（四）建立和完善基于智能场景的道德责任教育机制

The phenomenon of moral anomie arising from the deep embedding of artificial intelligence technology in ideological and political education is essentially a crisis of humanistic values caused by the usurpation of value rationality by technical rationality. Establishing a mechanism for moral responsibility education that is compatible with intelligent scenarios has become an effective way to resolve this problem. First, we must improve the moral responsibility education system suited to the digital-intelligent era. Implement an AI literacy education project, construct a moral responsibility cultivation model that conforms to the characteristics of the digital-intelligent era, and specifically set up courses on the ethics of AI responsibility to cultivate and strengthen students’ ethical awareness of responsibility in using AI. Use digital twin technology to conduct mirror training; through comparative analysis of moral decision-making simulations in virtual scenarios and real-world behavioral feedback, strengthen the cognition of the identity of virtual and real responsibilities, while simultaneously enhancing human-machine collaborative decision-making training. Design intelligent interactive scenarios with moral dilemma characteristics to cultivate students’ autonomous responsibility control ability under technical assistance. Relying on cutting-edge technologies such as virtual reality and augmented reality, develop immersive and highly realistic ideological and political education teaching systems based on multimodal technology, enabling educated subjects to experience moral dilemmas in lifelike intelligent interactive environments, and continuously strengthen the subject’s moral judgment and sense of responsibility through repeated practice. Second, construct a distributed responsibility network system. On the one hand, establish an algorithmic responsibility traceability system, embedding an ethical impact assessment module at the algorithm development stage to achieve full-process responsibility marking from code writing to educational application. On the other hand, construct a multi-subject responsibility-sharing mechanism and improve the moral responsibility assessment system. Use a distributed responsibility network system to clarify the responsibility boundaries of developers, educators, administrators, and learners, and accordingly develop multi-dimensional assessment indicators including technical ethics, educational ethics, and academic ethics, incorporating the assessment results into the education quality monitoring system. Third, establish a new integrity education mechanism based on intelligent scenarios. Give full play to the leading role of ideological and political education, effectively alleviate the integrity crisis from aspects such as ethical education, institutional norms, supervisory means, and cultural construction, cultivate students’ integrity awareness and sense of responsibility in using AI technology, and strengthen moral judgment and self-discipline abilities; improve scientific research norms, clarify the ethical boundaries of AI in knowledge production and application, and establish targeted punishment mechanisms for academic misconduct; enhance the ability to counter the misuse of intelligent technology with technology, develop and promote new types of academic misconduct detection technologies targeting AI; strengthen the construction of an integrity culture in intelligent education, and foster a new trend of integrity ethics education.

人工智能技术深度嵌入思想政治教育引发的道德失范现象，本质上是技术理性对价值理性的僭越所导致的人文价值危机，建立与智能场景相适配的道德责任教育机制成为破解这一难题的有效途径。一是健全适应数智时代的道德责任教育体系。实施人工智能素养教育工程，构建符合数智化时代特征的道德责任培养模式，专门设置人工智能责任伦理课程，培养和强化学生运用人工智能的责任伦理意识。运用数字孪生技术开展镜像训练，通过虚拟场景中的道德决策模拟与现实行为反馈的对照分析，强化虚拟和现实责任的同一性认知，同时增强人机协同决策训练，设计具有道德困境特征的智能交互场景，培养学生在技术辅助下的自主责任控制能力。依托虚拟现实、增强现实等前沿技术，开发基于多模态技术的沉浸式和高仿真的思想政治教育教学系统，使受教育者能够在逼真的智能交互环境中体验道德困境，通过反复实践不断强化主体的道德判断力和责任意识。二是构建分布式责任网络体系。一方面，建立算法责任溯源制度，在算法开发阶段嵌入伦理影响评估模块，实现从代码编写到教育应用的全流程责任标记。另一方面，构建多主体责任共担机制和完善道德责任评估体系，以分布式的责任网络体系明确开发者、教育者、管理者和学习者的责任边界，并按此开发包含技术伦理、教育伦理、学术伦理在内的多维度评估指标，将评估结果纳入教育质量监测体系。三是建立基于智能场景的新型诚信教育机制。充分发挥思想政治教育的引领性作用，从伦理教育、制度规范、监督手段和文化建设等方面对诚信危机进行有效纾解，培养学生使用人工智能技术的诚信意识和责任感，强化道德判断能力和自律能力；完善科研规范机制，明确人工智能在知识生产和应用中的伦理边界，建立有针对性的学术不端惩罚机制；提升以技术反制智能技术作恶的能力，开发针对人工智能的新式学术不端检测技术推广应用；加强智能教育的诚信文化建设，营造诚信伦理教育新风尚。

(The author is an associate professor at the School of Marxism, Beijing University of Technology.)

（作者系北京工业大学马克思主义学院副教授）

Online editor: Ma Jingren

网络编辑：马京仁