The Basic Logic and Contradiction Adjustment of Generative AI in Ideological and Political Education

Represented by GPT-4o, generative AI demonstrates core capabilities such as cross-modal generation, real-time interaction, and emotional expression, which are specific manifestations of the “intelligent emergence” of large AI models. Cross-modal generation refers to generating content of one modality based on information from another modality, such as text generating images, text generating videos, etc. Regardless of whether it is written text, visual images, video footage, or audio speech, advanced generative AI can skillfully understand, reason, convert, generate, and present these diverse information modalities and expressions, including language, charts, formulas, code, and music. Generative AI can also achieve real-time interaction, with GPT-4o having an average response time of only 320 milliseconds to voice input, which is close to human reaction time. In addition, GPT-4o can perceive user emotions through visual and audio information and respond accordingly, creating a more human-like interactive experience. However, at this stage, generative AI is not without its flaws. Due to systematic defects in data sources, training processes, and reasoning processes, large models may produce “hallucinations”; sometimes the generated content may contradict facts or fail to answer questions, and its values may not yet align with human values. Therefore, it is necessary to fully leverage the functional advantages of generative AI while correctly recognizing and addressing its technical limitations, steadily promoting the evolution and upgrading of elements in ideological and political education, and providing new opportunities for knowledge creation, dissemination innovation, and individual development.

(1) Knowledge Creation: Transforming Educational Content Production Models

Before the rise of generative AI, knowledge creation in the field of ideological and political education mainly followed two traditional models: Professional Generated Content (PGC) and User Generated Content (UGC). Experts, scholars, and ideological and political theory course (hereinafter referred to as “ideological and political courses”) teachers (collectively referred to as “educators”) mainly undertake the design, organization, and optimization of ideological and political education content, achieving the dissemination and infiltration of ideology and the enhancement of the ideological and political quality of the educated through the compilation of teaching materials, formulation of teaching plans, and implementation of education and teaching.

With the rapid development of internet technology, the boundaries of information dissemination and ideological exchange have been greatly expanded. Internet platforms have become a vast stage for the convergence of ideas and cultural collisions, where a large number of lively, vivid, and diverse opinions and cases have emerged in various media forms such as text, images, and videos, greatly enriching the volume, carriers, and discourse forms of ideological and political education content. However, the openness of the online environment has also brought challenges such as the difficulty of distinguishing the authenticity of information and the uneven quality of content, still requiring educators to carefully select and process massive amounts of information to extract valuable knowledge and transform it into effective educational content.

With the widespread application of generative AI, humans are no longer the only subjects of content production and knowledge creation. A new model—AI Generated Content (AIGC)—has emerged. Well-trained generative AI large models can automatically generate new content in various modalities such as text, images, videos, and audio based on user input prompts. AIGC is essentially a product of the deep integration of human experience and machine intelligence, where large models determine a vast number of model parameters by learning the inherent characteristics and organizational rules of training data, allowing them to quickly compute and output corresponding new content upon receiving new input information. In addition, generative AI can enhance user instructions, helping users better tame and utilize large models. For example, users can refine prompts through ChatGPT and then input them into models like DALLE-3 or Sora to generate higher quality and more accurately expressed images or video content.

The emergence of the AIGC model marks the formal intervention of machines in the knowledge creation process, with machines taking on tasks such as information collection, selection, refinement, integration, transformation, and initial creation. This effectively reduces the reliance of traditional content production models on human rich knowledge reserves, solid multimedia skills, and high aesthetic taste. Users only need to input natural language prompts and set simple parameters to quickly obtain content materials, even various exquisite and expressive digital works. This new paradigm of knowledge creation through human-intelligence collaboration will profoundly transform the production model of ideological and political education content, greatly improving production efficiency and providing a continuous supply of multimodal content resources for ideological and political education.

(2) Dissemination Innovation: Extending the Role of Educational Methods

Xi Jinping emphasized at the National Conference on Ideological and Political Work in Colleges and Universities: “We must use new media and new technologies to invigorate our work, promote the high integration of traditional advantages of ideological and political work with information technology, and enhance the sense of the times and attractiveness.” Generative AI technology can compress spatial and temporal distances through various media forms, breaking away from the previous static and singular narrative methods, and promote the extension of the role of traditional ideological and political education methods towards a three-dimensional direction, achieving profound innovation in dissemination.

From the horizontal expansion of method types, generative AI plays the role of an “intelligent learning companion”. It can not only provide information, impart knowledge, and guide creation, but also accurately match learning resources according to user needs, thereby meeting the needs of personalized development. For example, the “i Ideological and Political Large Model” at the University of Electronic Science and Technology of China includes ten nationally compiled textbooks for ideological and political education, over 100 classic works of Marxism-Leninism, and over 1000 articles from Marxist theory journals, which can be used for professional Q&A Tsinghua University has launched the “Qing Xiaoda” AI intelligent companion for learning growth assistant, which can automatically match and recommend quality educational resources based on students’ academic interests. In addition, generative AI can also serve as a “tool’s tool”, converting natural language into machine language to help educators quickly master data analysis tools, assist in developing software for teaching practice, and evaluate the appropriateness of various educational methods in different educational scenarios, guiding educators to select appropriate educational technologies and fully stimulate the application potential of advanced digital technology in the field of ideological and political education.

From the vertical deepening of method accessibility, generative AI can convey educational content through diverse carriers, enriching the experiential scenarios in the educational process, stimulating multiple senses of the educational subjects, and making it easier to achieve ideological resonance and value recognition. Cross-modal generative AI can construct a holographic environment, which can reproduce historical scenes through video generation, narrate through voice synthesis, and realize the “presence” experience of educational subjects through oral interaction, providing an immersive experience. The emotional perception and expression capabilities of generative AI can also provide emotional care for educational subjects.

From the deep extension of method fields, generative AI can promote the innovative development of the role of ideological and political education, conducive to deeply implementing the concept of “big ideological and political courses”, allowing educational subjects to receive ideological and political education in a broader virtual and real space, stimulating new vitality for collaborative education in society. For instance, Gannan Normal University has developed the “Red of the Soviet Area” ideological and political education large model based on the knowledge engine of revolutionary history in the Soviet area, helping students solidify their knowledge foundation and carry out social practice.

(3) Individual Development: Reshaping the Capability Matrix of Educational Subjects and Objects

The fundamental purpose of ideological and political education is to cultivate virtue and nurture people, promoting the comprehensive development of individuals. Educators and educational subjects are the core subjects and objects in ideological and political education. In the process of mutual engagement between humans and technology, technology drives changes in the capability matrix of subjects and objects, with the overall direction aligning with the ultimate goal of ideological and political education to promote individual development and self-transcendence.

Firstly, it hones the critical thinking of subjects and objects. Proper use of generative AI can encourage individuals to develop a habit of cautious thinking, laying a long-term cognitive foundation for comprehensive human development. On the one hand, generative AI opens up new perspectives for evaluating and improving teaching and learning outcomes, capturing gaps in the teaching process through data analysis and personalized evaluation, assisting educational subjects and objects in timely correcting deviations; on the other hand, subjects and objects need to possess discernment skills to judge the authenticity, accuracy, and effectiveness of the information output by generative AI, which also cultivates independent thinking and critical thinking, further enhancing the reflective capacity of educational subjects and objects.

Secondly, it shapes the global insight of subjects and objects. Proper use of generative AI can train individuals to establish a systemic view, building a solid cognitive framework for comprehensive human development. Through deep analysis and simulation prediction, generative AI can bridge the gap between the concrete and the abstract, guiding users from the local to the global, from the superficial to the essential. At the same time, integrating generative AI into ideological and political education practice requires subjects and objects to accurately grasp the essence of problems, select tools reasonably, and ensure that the output content meets expectations through step-by-step analysis and guidance, facilitating the continuous improvement of the personal cognitive system of subjects and objects and deepening their insight into the whole picture.

Thirdly, it stimulates the subjective initiative of subjects and objects. Proper use of generative AI can motivate individuals to form a proactive awareness, injecting spiritual energy into comprehensive human development. Generative AI lacks initiative and requires users to input instructions to trigger it, and it is often difficult to provide the expected responses when facing vague problems or partial descriptions; coupled with the rapid iteration of technology and the scarcity of mature experience, users often need to explore and navigate through practice. Therefore, educational subjects and objects must take the initiative, consciously understand the principles of technology, algorithm logic, and cutting-edge developments, and maintain a high sensitivity to the characteristics, advantages, and limitations of various generative AI products. Through continuous practice, they can optimize adaptation strategies to truly achieve the maximization of value through technology mastery.

“Contradictions exist in the development process of all things and run through the entire process of every thing’s development.” The ideological and political education process itself contains various contradictions, the most significant of which is the contradiction between the expected requirements of people’s ideological and political quality under certain social conditions and the actual situation. Generative AI is expected to become an important external factor in promoting the transformation of contradictions in ideological and political education, but it is still immature in technology, application, and governance, and may also trigger new contradictions.

(1) The Contradiction Between Content Generation Uncertainty and Value Transmission Certainty

Ideological and political education led by the Communist Party of China has always been guided by Marxism, deeply rooted in the rich theoretical and practical soil of socialism with Chinese characteristics, while also drawing nourishment from the profound heritage of excellent traditional Chinese culture. In cultivating socialist builders and successors, it maintains a firm stance and a clear banner without wavering. However, generative AI, based on a big data-driven and probabilistic recombination operating mechanism, may produce biased information, vague viewpoints, or even non-mainstream opinions, posing significant risks in the ideological field. This contradiction exists between the uncertainty of content generation and the certainty requirements of value transmission in ideological and political education.

The quality of the content output by generative AI and the values it embodies largely depend on the data and algorithm logic it has learned, similar to the “Chinese Room” thought experiment. On the one hand, it is difficult to manually proofread and screen massive amounts of learning data one by one, and there is a lack of effective data credibility assessment components during the model training phase; when introducing online information as a dynamic updating data source, the quality control of learning data becomes even more challenging; coupled with the fact that the algorithm itself may amplify biases in the data, personal values and understanding of things may directly reflect in the output results, leading to biases in conclusions due to imbalanced training data. On the other hand, it is currently common to incorporate user feedback into a closed loop to feed back into the model to optimize output content, but this may also lead generative AI to tend to produce results that meet user expectations, thus under specific manipulations, the preferences of a minority group may be widely disseminated, further exacerbating the possibility of “distortion” in conclusions and trapping users in an “information cocoon” dilemma.

The ideological risks posed by the uncertainty of content generation must be taken seriously in the context of ideological and political education. Brazilian scholars have conducted simulated experimental studies proving that ChatGPT exhibits strong and systematic political bias, with a noticeably “left-leaning” political stance. Recent research by the American tech company OpenAI also shows that ChatGPT adjusts its responses based on inferred gender and race from the username, displaying certain “stereotypes”. Moreover, generative AI sometimes also exhibits ambiguous stances; for example, the movie “Pacific Rim” is a typical case of the United States promoting American-style heroism and implementing “invisible propaganda”. However, when asked about the American values embodied in this film on a certain domestic generative AI product, its response positively mentioned individual heroism, teamwork, collectivism, globalization, and multiculturalism, which raises suspicions of “avoiding the heavy and focusing on the light”. Such uncertain and uncontrollable information can quietly permeate into the cognition, thoughts, and value formation processes of adolescents. When adolescents become confused about mainstream ideology and seek answers from generative AI, if generative AI fails to clearly, deeply, and thoroughly explain the principles, or even in extreme cases provides biased information, deviating from the mainstream discourse system, it may deepen misunderstandings and gaps, leading to a departure from the original intention of ideological and political education.

(2) The Contradiction Between Insufficient Supply of Vertical Large Models and the Professional Needs of Ideological and Political Education

Vertical large models are customized large models targeted at specific fields or industries, incorporating a wealth of theoretical knowledge and practical cases from that field or industry during training, thus possessing high professionalism and adaptability. Ideological and political education, as a systematic discipline, encompasses rich content including ideological education, political education, moral education, and legal education, and is logically coherent and self-contained. However, general large models lack targeted content supply, leading to insufficient competence. Currently, there is no widely applicable ideological and political education large model in the education sector, resulting in a contradiction between insufficient model supply and the professional needs of ideological and political education.

The insufficient supply of vertical large models applicable to ideological and political education is, in fact, a result of social choices. When ChatGPT was first released, general large models were the most popular. From ChatGPT to GPT-4o, they represent general large models. The degree of generalization marks the level of development of generative AI, and artificial general intelligence (AGI) is recognized as the next goal of AI development. Many tech giants have rushed into this field, competing for technological supremacy, hoping to gain an advantage in the next leap of AI. However, the climb to technological heights is fraught with difficulties, involving multiple games of capital, technology, industry, and even international politics. The market has gradually shifted towards the development of vertical large models in high-value-added industries, with various industries laying out plans to assist enterprises in reducing costs and increasing efficiency through the application of vertical large models, creating considerable economic benefits. However, since ideological and political education does not have a direct commercial nature, it faces a natural disadvantage in this large model competition. Although the Ministry of Education has begun to build ideological and political education large models, universities such as Tsinghua University and Shanghai Jiao Tong University have initiated explorations into applying generative AI to ideological and political education, and universities like the University of Electronic Science and Technology of China and Gannan Normal University have developed ideological and political education models and achieved initial results, the current coverage of educational content and the beneficiaries of educational subjects are relatively limited, and a widely applicable ideological and political education large model has not yet been established, nor has a large-scale and systematic educational synergy been formed.

Currently, due to the insufficient supply of large models, the depth of application of generative AI in ideological and political education is limited. Taking text-based generative AI as an example, although it performs well in general functions such as information retrieval and writing assistance, when educators attempt to use it for course preparation, automatic question generation, or assignment grading, or when educational subjects expect it to deeply interpret textbook knowledge points and supplement richer arguments, general large models often fall short. Due to the lack of relevant corpus in the training data in the field of ideological and political education, when faced with problems that have strong disciplinary professionalism and have not been covered in training, general large models tend to be superficial and vague, though correct, but not necessarily aligned with the discourse system of ideological and political education, lacking specificity, precision, and depth. While general large models can play the role of a “learning companion”, they struggle to fulfill the responsibilities of ideological and political course teachers.

(3) The Contradiction Between Advanced Technology Empowerment and Educational Application Risks

The potential of generative AI for development and its strong driving force for industrial innovation have been recognized by all sectors of society, yet the risks accompanying its intervention in ideological and political education cannot be underestimated. The education sector has always maintained close attention and a cautious attitude towards its application; there exists a contradiction between the empowerment of generative AI technology for innovative development in ideological and political education and the risks it may trigger.

Firstly, there are cumulative risks. As a specific application scenario for generative AI, ideological and political education is positioned at the downstream application layer in the industrial chain. The full-process development of generative AI requires high capital and technical demands, and downstream applications often regard large models as a type of infrastructure, deploying them quickly and at low cost by calling interfaces and feeding professional data. However, as the industrial chain extends, the risks of errors and loss accumulate like a snowball, with unknown potential risks in the training data quality, computing performance, and stability at the foundational layer, as well as in the algorithm’s credibility and feedback optimization mechanisms at the model layer. These risks will be transmitted downstream, while the customization development process at the application layer will introduce new corpus and adaptive adjustments, which will also bring in risk factors, compounding the aforementioned risks and forming greater “cumulative risks”.

Secondly, there are ethical risks. Since the release of ChatGPT, there have been multiple incidents of commercial data leakage, copyright infringement, illicit information acquisition, and the spread of online rumors. In the field of ideological and political education, sensitive ethical issues involve academic integrity and personal data rights. For example, when educational subjects use generative AI assistants like ChatGPT to complete assignments or papers, the boundary between plagiarism and reasonable reference becomes blurred; furthermore, using large models like GPT-4o with emotional perception capabilities to monitor the emotional states of educational subjects in real time through visual and auditory media may help educators flexibly adjust teaching methods, but this may raise disputes regarding personal privacy and data security. These emerging issues pose severe challenges to existing social ethics and legal regulations, and as generative AI continues to develop rapidly, new functions, new scenarios, and new applications are constantly emerging, making it likely that new problems will continue to arise, exacerbating ethical risks and heightening public concerns.

Thirdly, there are imbalance risks. Technological rationality and value rationality are a pair of relationships that ideological and political education must handle well in the process of digital transformation. Technological rationality pursues the rationality, normativity, and effectiveness of technology, aiming to enhance productivity and drive economic and social development; value rationality emphasizes a people-centered approach, achieving the free and comprehensive development of individuals and sustainable social development, representing the inherent requirements for technology to serve good. The two should maintain a balance, but under the dominance of technological supremacism and the spontaneous impact of the market, technological rationality often prevails, squeezing the living space of value rationality. This imbalance specifically manifests in the field of ideological and political education as an excessive emphasis on artificial intelligence topics, overestimating the effectiveness of generative AI in ideological and political education, and viewing intelligence as a panacea for solving difficult problems. The decline of value rationality will lead to “formalism” and “technicism”, resulting in negative impacts that invert the essence and the means, and the strong entry of generative AI exacerbates this “imbalance risk”.

In the face of the specific contradictions arising from generative AI’s intervention in ideological and political education, it is urgent to explore effective adjustment paths, proactively create favorable conditions, and maximize the positive empowering effect of generative AI on ideological and political education.

(1) Aligning Value Concepts to Build a Solid Foundation for Technological Empowerment

To address the contradiction between content generation uncertainty and value transmission certainty, it is necessary to guide the alignment of large model values, enhance the supply of ideological and political education content, and optimize algorithm mechanism design to build a solid foundation for generative AI’s empowerment of ideological and political education.

Guiding the alignment of large model values aims to ensure that the values of large models are consistent with the goals, concepts, and standards of ideological and political education. The “Interim Measures for the Management of Generative Artificial Intelligence Services” clearly stipulates that any generative AI service provided to the public within China must adhere to socialist core values and must take effective measures to prevent discrimination based on ethnicity, belief, nationality, region, gender, age, occupation, health, etc. Based on this, ideological and political educators should actively participate in the research on aligning large model values, playing a role in defining value connotations, controlling discourse scales, and evaluating rule effectiveness, providing support for achieving diverse and comprehensive value alignment. In daily use, educators should also actively engage in training large models through human-machine dialogue, taking practical actions to maintain balance in the discourse field, safeguard ideological security, and ensure fairness and impartiality.

Enhancing the supply of ideological and political education content means systematically and at scale embedding ideological and political education content into training data, ensuring that mainstream value concepts in China run through the underlying data sources. Ideological and political educators should lead the construction of an ideological and political education resource library, which should include theoretical guidelines such as Marxist theory and the theoretical system of socialism with Chinese characteristics, as well as vivid stories depicting the remarkable achievements of socialism with Chinese characteristics, and carry ideological consensus from excellent traditional Chinese culture, revolutionary culture, and advanced socialist culture, thereby strengthening the digital construction of the discourse system and teaching generative AI to articulate principles profoundly, thoroughly, and vividly. In addition, it is necessary to continuously update and enrich the resource library to ensure that the discourse of ideological and political education keeps pace with the times.

Optimizing algorithm mechanism design requires designing effective correction and monitoring mechanisms during the model construction, training, and reasoning stages to avoid the intrusion and output of misleading information, erroneous content, and non-mainstream discourse. Developers should consciously regard value alignment as a core principle in the development of large models, using techniques such as bias correction, error correction, and data enhancement to reduce the impact of data biases, and regularly review and update models to ensure that mainstream value concepts remain dominant. At the same time, it is necessary to strengthen the supervision of generated content, setting up interception management modules during the reasoning phase, ensuring that even content that deviates from mainstream values and ideological stances is prevented from being output to users, using technical firewalls to help anchor correct value orientations.

(2) Improving the Technical Ecosystem to Create a Positive Environment for Technological Empowerment

The demand for specialized large models in ideological and political education and their potential application risks are systematic issues that require systematic solutions. The primary task is to continuously improve the technical ecosystem, focusing on technological breakthroughs while enhancing ethical constraints, and widely conducting public education to create a positive environment for generative AI’s empowerment of ideological and political education.

Firstly, promote technological breakthroughs and service innovations to improve the internal ecosystem of generative AI technology, which serves as the source of empowerment. Specialized ideological and political education large models should be integrated innovation platforms, collaboratively planned by ideological and political educators, managers, and technology developers under the coordination of the Ministry of Education, starting from the core needs of ideological and political education to scientifically plan the functional map and define application boundaries. This can not only free ideological and political education from technological constraints, facilitating its digital transformation but also effectively integrate resources to avoid fragmentation of application effects due to resource dispersion. In addition, to meet the diverse needs of daily education and teaching, this platform should also open customization windows, encouraging educators to customize specific ideological and political education sub-models based on specific teaching goals, promoting the widespread implementation of new forms of intelligent and virtual ideological and political education. At the same time, attention should be paid to the safety of ideological and political education large models, striving to achieve full-process controllability from computing resources, algorithm logic, to data management, building a solid network security defense line to resist external value shocks and encroachments on public opinion.

Secondly, strengthen ethical constraints and legal regulations to improve the external institutional ecosystem of generative AI technology, which is the foundation of trust for technological empowerment. Educators should take the initiative to anticipate technical application risks, particularly those related to ideological security, intellectual property protection, academic integrity, data security, and personal privacy, and actively collaborate with other departments to explore risk prevention and governance methods. Relevant government departments should also cooperate, continuously refining and improving related laws and regulations based on compliance with the “Interim Measures for the Management of Generative Artificial Intelligence Services”, clarifying responsible parties, and strengthening technical safety supervision; considering the particularity of ideological and political education, more specific and feasible technical specifications and ethical guidelines should be formulated to alleviate doubts and controversies in practical applications.

Thirdly, promote public education, disseminate rational cognition, and optimize the external social environment for generative AI technology to ensure stable and sustainable technological empowerment. The strong emergence of generative AI has sparked widespread attention and strong reactions from all sectors of society, including negative concerns, such as the phenomenon of “structured abandonment” caused by technological divides, the “systemic refugee” phenomenon trapped in disorderly human-machine interactions, and the widespread anxiety about “technological unemployment”. In the face of the turbulent social mentality, educators should take on the responsibility of ideological and political education, playing the roles of spiritual motivation, action leadership, and wisdom enlightenment, addressing social livelihood issues triggered by rapid technological development, guiding the public to rationally review technological changes through educational outreach, online discourse, and scientific research breakthroughs, promoting the compliant, reasonable, and empathetic application of generative AI products, actively adapting to the profound changes in technological patterns and social structures, while coordinating various forces to proactively plan and respond to challenges, jointly promoting inclusive economic and social growth.

(3) Enhancing Digital Literacy to Embrace a Bright Future Empowered by Technology

Whether objectively responding to the inherent limitations of current generative AI content uncertainty, accelerating the bridging of the practical gap between the supply and demand of specialized large models in ideological and political education, or cautiously addressing potential educational application risks, it is essential to stimulate the subjective awareness of individuals. Educational subjects and objects should actively enhance their information discernment and correction capabilities, consciously improve their ability to use intelligent technologies, and strive to shape leading capabilities for a digital future, stabilizing through the “growing pains” of the collision and integration of generative AI and ideological and political education, and embracing the bright future of generative AI empowering ideological and political education.

It is urgent to deepen the understanding of technological boundaries and enhance the sensitivity to information discernment and correction. While generative AI demonstrates innovative potential in discourse expression, dissemination media, and emotional communication, it currently cannot guarantee that the output content is completely correct and objectively neutral. Therefore, educational subjects and objects must continuously arm themselves with scientific theories when using generative AI, deeply grasp the inherent laws of content production, guarding both the entry to prevent sensitive information leakage and the exit to accurately analyze and correct erroneous thoughts. Recognizing the “capabilities” and “incapabilities” of technology, when faced with generated content that exceeds personal cognitive scope, individuals should not blindly trust it but should utilize search engines and other means for cross-validation to break free from the confines of the “information cocoon”.

In the long run, it is also necessary to cultivate a lifelong learning mindset and enhance the ability to master intelligent technologies. In the past decade, the rapid iteration of digital technologies has accelerated, and both educational subjects and objects should establish a lifelong learning mindset, striving to adapt to new situations and trends. Educators should actively participate in discussions and professional training on the cutting edge of technology with an open attitude, continuously enhancing their ability to deeply integrate intelligent technologies with education and teaching; actively exploring the application of intelligent technologies in teaching practices, investigating effective paths for enhancing personal comprehensive quality under human-intelligence interaction modes. Educators should also courageously attempt interdisciplinary collaboration, strengthening communication and exchange with experts and scholars in fields such as information technology and law, to build a more comprehensive technological cognition framework for more effectively advancing technological practices. For educational subjects, they should also actively adapt to the developments and changes in intelligent technologies, fully utilizing online resources for self-directed learning, striving to seize opportunities, and appropriately applying intelligent technologies in their academic studies to promote personal growth.

Looking to the future, it is essential to strengthen digital literacy education, forging the leading force for a digital future. Digital literacy, as a broad concept, encompasses individuals’ survival abilities, thinking abilities, production abilities, and innovation abilities in a digital society, as well as ethical and moral qualities related to national security, social stability, and personal rights, making it an indispensable core quality in the digital age. Digital literacy education and ideological and political education both aim to cultivate well-rounded individuals; the former cultivates talents needed for the modernization process, while the latter provides value guidance and strategic direction for the former. Therefore, cultivating digital literacy is crucial for educational subjects and objects. Specifically, educational subjects should systematically learn digital knowledge, skills, and ethical standards through classroom teaching and online training to enhance social responsibility; while educators should not only be adept at using generative AI tools but also stand at the forefront of technological development, drawing a blueprint for the innovative development of ideological and political education empowered by digital technologies, continuously optimizing human-intelligence collaboration models from the demand side, and achieving a transition from followers to leaders in this field.