Algorithmic Hegemony and Ideological Manipulation: An Exploration of the Path by Which Western States Discipline Algorithms

In the context of the digital age, intelligent algorithms have transformed into a key technological force shaping the landscape of information dissemination and the construction of social cognition. Western states, relying on their first-mover advantage in the field of science and technology, meticulously and systematically discipline intelligent algorithms, skillfully coupling them deeply with their own ideology, thereby constructing a highly influential system of algorithmic hegemony. At the same time, through a series of carefully designed laws, regulations, industry standards, and policy guidance mechanisms, Western states discipline the development, application, and oversight processes of intelligent algorithms in all respects, ensuring that algorithms, in the course of their operation, conform to their established ideological orientation, thus becoming a powerful yet concealed tool for promoting their own ideology. This gives rise to a global, deep-seated game concerning algorithmic sovereignty and ideological security, the impact of which is far-reaching and complex, worthy of in-depth exploration and analysis.

在数字化时代背景下，智能算法已然蜕变为左右信息传播格局与社会认知建构的关键性技术力量。西方国家倚仗其在科技领域的先发优势，对智能算法进行精心且系统化的规训，巧妙地将其与自身意识形态深度耦合，从而构筑起一套极具影响力的算法霸权体系。与此同时，西方国家借助一系列精心设计的法律法规、行业标准以及政策引导机制，全方位地规训着智能算法的开发、应用与监管流程，确保算法在运行过程中能够契合其既定的意识形态导向，进而成为其推行自身意识形态的有力且隐匿的工具。由此引发关乎算法主权与意识形态安全的全球性深度博弈，其影响深远且复杂，值得深入探究与剖析。

I. Control Mechanisms: The Panoramic Infiltration and Intervention of Western Algorithmic Power in Social Life

一、控制机制：西方算法权力对社会生活的全景式渗透与干预

The rise of intelligent algorithms marks the birth of a new form of social control. During the period of industrial capitalism, social control manifested as what Michel Foucault termed the “disciplinary society,” where discipline and standardization were the core means of achieving control, implemented through physical institutions such as factories, prisons, hospitals, and schools, forming a model of “panopticism.” In contrast, the centralization of algorithmic platforms has given rise to a new form of social control, one that echoes the concept of the “society of control” discussed by Gilles Deleuze. In the “society of control,” control is achieved not through fixed disciplinary institutions, but through dynamic, adaptable algorithmic mechanisms that can flexibly shape and guide individual behavior without relying on traditional disciplinary means. Western states, utilizing algorithmic hegemony, not only comprehensively monitor individual lives but also exert influence on other countries, raising serious concerns about privacy, data protection, and national sovereignty.

智能算法的兴起标志着一种新型社会控制形态的诞生。在工业资本主义时期，社会控制体现为米歇尔·福柯（Michel Foucault）所称的“规训社会”,其中纪律和标准化是实现控制的核心手段，通过实体机构如工厂、监狱、医院和学校实施控制，形成“全景敞视主义”模式。相较之下，算法平台的中心化催生了一种新的社会控制形态，这一形态与吉尔·德勒兹（Gilles Deleuze）所讲的“控制的社会”理念遥相呼应。在“控制社会”中，控制不是通过固定的规训机构，而是通过动态的、适应性强的算法机制来实现，这些机制能够灵活地塑造和引导个体行为，无需依赖传统的规训手段。西方国家利用算法霸权，不仅全面监控个人生活，还对他国施加影响，引发了人们对隐私、数据保护和国家主权的严重关切。

(1) Deprivation of Privacy: Implementing Algorithmic Surveillance of Global Citizens

（一）剥夺隐私：实施对全球公民的算法监控

The rapid development of intelligent algorithm technology has led us into a new era beyond the traditional “panopticon” model—the era of the “post-panopticon.” This shift means that surveillance is no longer limited to the disciplining of individuals but has expanded to the collection, analysis, and prediction of data. This concept of “liquid” surveillance echoes the theories of Zygmunt Bauman. Bauman argued that in the context of digitalization and economic globalization, surveillance has become more fluid, flexible, and almost invisible, permeating every corner of society like a liquid, which he termed the “liquid modern world.” Liquid surveillance no longer requires fixed watchtowers or walls; instead, through ubiquitous smart devices and algorithms, it conducts real-time monitoring of individuals’ behaviors, habits, and preferences. The purpose of liquid surveillance is not merely discipline and punishment; more importantly, it is to collect data, analyze behavioral patterns, and predict future trends. Western states, leveraging the power of algorithms and data giants, have established a technical system for the comprehensive surveillance of the global population through the collection and analysis of internet data via algorithms. For example, in the early 21st century, the U.S. Department of Defense commissioned Acxiom Corporation to assist in building a surveillance system, and the “PRISM” program exposed by Edward Snowden in 2014 [1]1 The PRISM program was a clandestine mass electronic surveillance data mining program operated by the U.S. National Security Agency (NSA) from 2007, revealed by whistleblower Edward Snowden in 2013.—these means covered almost all online behaviors of global internet users, from email content and web browsing history to online chat records, all of which were difficult to escape surveillance, with vast amounts of data flowing into U.S. data systems.

智能算法技术的迅猛发展引领我们进入了一个超越传统“全景监狱”模式的新时代——“后全景监狱”时代。这一转变意味着，监视不再局限于对个体的规训，而是扩展到了数据的收集、分析和预测。这种“液态”监视的理念与齐格蒙·鲍曼（Zygmunt Bauman）的理论相呼应。鲍曼认为，在数字化和经济全球化的背景下，监视变得更加流动、灵活，几乎无形，就像液体一样渗透到社会的每一个角落，他称之为“液态的现代世界”。液态监视不再需要固定的监视塔或围墙，而是通过无处不在的智能设备和算法，对个人的行为、习惯和偏好进行实时监控。液态监视的目的并不仅仅是为了规训和惩罚，更重要的是为了收集数据，分析行为模式，并预测未来趋势。西方国家借助算法和数据巨头的力量，通过算法对互联网数据的收集和分析，从而建立了全方位监视全球人口的技术系统。如21世纪初美国国防部委托安克诚公司协助建立监视系统，2014年斯诺登曝光的“棱镜计划”,这些手段几乎覆盖了全球网民的所有网上行为，从电子邮件内容、网页浏览记录到在线聊天记录等全部难逃监视，大量数据流向美方的数据系统。

Within the super “post-panopticon” of Western algorithmic discipline, the behaviors, thoughts, emotions, personalities, and social relationships of global citizens have become transparent. All actions of citizens in the online world are easily captured and recorded by algorithms, and through analysis, become data tags for constructing personal profiles, achieving comprehensive, round-the-clock surveillance of citizens. Research by Swedish researcher David Sumpter found that under the monopoly of Western media, the social lives of the global public have fallen into a state of near-naked transparency. Taking Meta (formerly Facebook) as an example, its algorithms, by analyzing user interaction behaviors such as likes, shares, and comments, can reveal user personality traits and preferences, sometimes even knowing users better than they know themselves. Meta’s algorithm requires only 15 posts for in-depth behavioral analysis; every click, browse, and interaction by users transmits information about their personality traits and emotional states. With technological progress, social platforms like Facebook may track and exploit user emotional responses more accurately, thereby achieving comprehensive surveillance of global audience groups. Western tech giants, relying on their possession of the world’s most advanced algorithmic technologies and vast data resources, have formed strong market influence and control through the widespread application of algorithms in areas such as search engines, social media, and e-commerce, making it difficult for other countries to compete.

在西方算法规训的超级“后全景监狱”中，全球民众的行为、思想、情感、性格、社会关系等都变得透明。民众在网络世界的一切行为，都会被算法轻易捕捉记录，并通过分析成为构建个人画像的数据标签，实现对民众全方位、全天候的监控。瑞典研究者大卫·萨普特（David Sumpter）的研究发现，在西方媒体的垄断下，全球公众的社会生活陷入了近乎赤裸透明的状态。以Meta（原Facebook）为例，其算法通过分析用户互动行为，如点赞、分享、评论等，能揭示用户性格特征和偏好，有时甚至比用户自身更了解自身。Meta的算法仅需15个帖子就能进行深入行为分析，用户的每一次点击、浏览和互动都在传递个性特征和情感状态的信息。随着技术进步，像Facebook这样的社交平台可能会更精准地追踪和利用用户情绪反应，从而实现对全球受众群体的全方位监控。西方科技巨头凭借其掌握的全球最先进的算法技术和庞大数据资源，通过算法在搜索引擎、社交媒体、电子商务等领域的广泛应用，形成了强大市场影响力和控制力，致使其他国家难以抗衡。

In this regard, Harvard researcher Shoshana Zuboff profoundly points out that Western society has irreversibly fallen into the vortex of “surveillance capitalism.” The core mechanism of surveillance capitalism is the transformation of personal information and privacy from a category of individual rights into a resource for creating economic value. This transformation marks a shift from industrial capitalism, centered on industrial production, to information capitalism, centered on information and data. Zuboff particularly emphasizes the concept of “instrumentarianism,” as a new form of power, “which no longer uses weapons and armies, but realizes its will through the automated medium of an increasingly ubiquitous computing architecture composed of ‘smart’ networked devices, things, and spaces.” The unlimited expansion of surveillance capitalism not only threatens the freedom and privacy of the global public but also leads to the continuous concentration of wealth, knowledge, and power in the West. Capitalist countries like the United States, using their mastery of core digital surveillance technologies, have constructed the world’s largest surveillance network and extended surveillance globally under the pretext of security. At the political, economic, and ideological levels, digital surveillance has become a tool of hegemony. Western states, in collusion with surveillance giants, suppress other economic entities, control the global economic lifeline, and fabricate the false ideology that “more surveillance equals a stronger sense of security,” making the global public submit to their hegemonic expansion.

对此，哈佛大学研究者肖莎娜·祖博夫（Shoshana Zuboff）深刻地指出，西方社会已经不可逆转地陷入了“监控资本主义”的旋涡。监控资本主义的核心机制是将个人信息和隐私从个人权利的范畴转变为创造经济价值的资源。这种转变标志着资本主义从以工业生产为中心的工业资本主义，向以信息和数据为核心的信息资本主义的转型。祖博夫特别强调了“工具主义”这一概念，作为一种新的权力形式，“它不再使用武器和军队，而是通过由‘智能’网络设备、事物和空间组成的日益无处不在的计算架构这一自动化媒介来实现自己的意志”。监控资本主义的无限扩张，不仅威胁全球公众自由与隐私，而且导致财富、知识和权力不断向西方集中。美国等资本主义国家利用其掌握的数字监控核心技术，构建了全球最大监控网络，并以安全为由将监控延伸至全球。在政治、经济和意识形态层面，数字监控已成为霸权工具，西方国家联合监控巨头打压其他经济实体，掌控全球经济命脉，制造“监控越多-安全感越强”的虚假意识形态，使全球民众服从其霸权扩张。

(2) Algorithmic Cage: Intelligently Recommending Information Content Aligned with Western Ideology

（二）算法牢笼：智能推荐符合西方意识形态的信息内容

Through personalized recommendation systems, Western states, based on the results of user data analysis, filter and push information content that aligns with their values and interests. This mechanism of information filtering and pushing leads users, when using platforms such as social media and news websites, to often only encounter content consistent with Western ideology, while other voices are marginalized or blocked. Intelligent recommendation systems use advanced algorithms to analyze user behavioral data, preferences, and social networks, predicting and pushing content that users may be interested in. Although this personalized service seems attractive, saving users time in filtering information, it also conceals issues of information control. Algorithms, like an invisible weaver, tailor-make for us an information siege—an “algorithmic cage” composed of intelligent recommendations.

借助个性化推荐系统，西方国家依据用户数据分析结果，筛选并推送符合其价值观和利益导向的信息内容。这种信息筛选与推送机制，导致用户在使用社交媒体、新闻网站等平台时，往往只能接触到符合西方意识形态的内容，而其他声音则被边缘化或屏蔽。智能推荐系统利用先进的算法分析用户的行为数据、偏好和社交网络，预测并推送用户可能感兴趣的内容。这种个性化服务虽看似诱人，节省用户筛选信息的时间，但其背后也暗藏信息控制问题。算法宛如一位隐形编织者，为我们量身定制了一个信息围城——一个由智能推荐构成的“算法牢笼”。

The “algorithmic cage” silently restricts the horizons of Western citizens, trapping them in a cocoon meticulously woven by algorithms, which becomes a breeding ground for the “filter bubble” effect and group polarization in online information. Algorithmic recommendation systems, through detailed analysis of user behavior and preferences—such as historical browsing records, search history, click behavior, and preference settings—filter and push content aligned with users’ views, thereby forming an information “bubble” around them. While the formation of this filter bubble can provide a more personalized user experience, sparing users from vast amounts of irrelevant information in the age of information explosion, it isolates users from differing or opposing opinions, limiting their exposure to diverse information. Similar viewpoints are continuously reinforced by algorithms, while dissenting voices are marginalized. Cass R. Sunstein argues that once a large degree of information homogenization occurs, “group members initially have certain inclinations; after deliberation, people continue to move in the direction of those inclinations, ultimately forming extreme views.” [2]2 Cass R. Sunstein, *Republic.com 2.0* (Princeton: Princeton University Press, 2007), p. 60. Due to its anonymity, decentralization, and tendency to form homogeneous groups, the online environment makes group polarization more pronounced than in real life. Intelligent algorithms constantly provide users with self-reinforcing information bubbles, reducing their opportunities to encounter and consider different perspectives. Consequently, some Western users can only access information and opinions close to their own cognition, turning a deaf ear to external dissenting voices, thereby exacerbating social division and polarization. For example, the BBC, when reporting on China, often uses a “hellish filter” [3]3 “Hellish filter” (阴间滤镜) refers to the practice of using dark, desaturated, or gloomy visual tones in media coverage to create a negative impression of a subject, often applied by Western outlets when reporting on China. to render originally bright and festive scenes dim and outdated, thereby shaping a negative image of China in the minds of Western citizens. This practice not only restricts Western citizens’ horizons but also intensifies information homogenization, intertwining the filter bubble effect with group polarization and further reinforcing Western citizens’ stereotypes of China.

“算法牢笼”在无声无息中限制了西方民众的视野，使其陷入了一个由算法精心编织的茧房，成为网络信息“茧房”效应和群体极化效应的温床。算法推荐系统通过细致分析用户的行为和偏好，如历史浏览记录、搜索历史、点击行为和偏好设置，为用户筛选并推送与其观点相关的内容，从而在用户周围形成一个信息的“气泡”。这种过滤气泡的形成，虽然可以提供更加个性化的用户体验，让用户在信息爆炸的时代中不必面对海量无关信息，但过滤气泡将用户与不同或相反的意见隔离开来，限制了他们接触到多样化信息的可能性，相似的观点在算法的推动下不断被加强，而不同的声音则被边缘化。凯斯·桑斯坦（Cass R. Sunstein）认为，一旦出现大量的信息同质化现象，“团体成员一开始即有某些偏向，在商议后，人们朝偏向的方向维续移动，最后形成极端的观点”。网络环境由于其匿名性、去中心化和易于形成同质群体的特点，使得群体极化现象在网络中比现实生活中更为显著。智能算法不断地为用户提供自我加强的信息泡沫，使其减少了接触和考虑不同观点的机会。一些西方用户因此只能接触到与自己认知相近的信息和看法，而对外界不同的声音充耳不闻，进而加剧了社会的分裂和极化。以BBC为例，其在报道中国时，常常使用“阴间滤镜”,让原本明亮、喜庆的场面变得晦暗、陈旧，从而在西方民众心中塑造中国负面形象。这种做法不仅限制了西方民众的视野，还加剧了信息同质化现象，使得信息茧房效应和群体极化效应相互交织，进一步强化了西方民众对中国的刻板印象。

(3) Political Intervention: Data Colonialism by Western Tech Giants

（三）政治干预：西方科技巨头的数据殖民主义

Western tech giants, through data extraction and algorithmic technology, have achieved manipulation of global political processes, a behavior that can be regarded as a “new form of colonialism.” Ulises A. Mejias, a scholar at the State University of New York at Oswego, and Nick Couldry, a scholar at the London School of Economics and Political Science, explore this phenomenon in detail in their work *The Costs of Connection: How Data Is Colonizing Human Life and Appropriating It for Capitalism*. They point out that large tech companies such as Meta, Amazon, and Alphabet collect information through activities like checking emails, counting steps, and online shopping, accumulating vast data resources. This data accumulation is not an accidental phenomenon of rapid growth in the tech industry but resembles the predatory behavior of historical colonialism. Data colonialism has three notable characteristics: first, globality—data extraction is global, with no place of human life immune to this form of exploitation. Second, enormous scale—large tech companies create unprecedented wealth through data extraction. Finally, it perpetuates dispossession and injustice—data extraction continues the dispossession and injustice that began with colonialism. By designing discriminatory algorithms and other means, tech companies convert data into wealth and power for themselves, rather than for users.

西方科技巨头通过数据攫取和算法技术，实现了对全球政治过程的操控，这种行为可以被视为一种“新的殖民主义形式”。美国纽约州立大学奥斯威戈分校的学者乌利塞斯·A. 梅加斯（Ulises A. Mejias）和伦敦政治经济学院学者尼克·库尔德利（Nick Couldry）在著作《数据攫取：科技巨头的新殖民主义及反击方法》中，详细探讨了这一现象。他们指出，Meta、亚马逊和Alphabet等大型科技公司通过人们查看电子邮件、计算步数、网上购物等行为收集信息，积累了庞大的数据资源。这种数据积累并非科技行业快速增长带来的偶然现象，而是类似于历史上殖民主义的掠夺行为。数据殖民主义具有三个显著特征：首先是全球性，数据攫取是全球性的，没有哪个地方的人类生活能够免受这种形式的剥削。其次是规模巨大，大型科技公司通过数据攫取创造了前所未有的财富。最后，延续了剥夺和不公正，数据攫取延续了殖民主义开始的剥夺和不公正。通过设计歧视性算法等手段，科技公司将数据转化为财富和权力，而不是用户的财富和权力。

Data colonialism has a profound impact on global political activities. In political campaigns, algorithms are used to analyze voter data, formulate precise campaign strategies and propaganda content, and even create and spread political rumors to influence voter decisions. This reliance makes government decision-making more dependent on tech companies’ data and technology, further strengthening the influence of tech giants and thereby consolidating their neocolonial control mechanisms on a global scale. Data colonialism also has a profound impact on social structures and economic inequality. Through a new type of human relationship—data relations—data can be extracted and commodified. Social life worldwide has become an “open” resource available for extraction, accessible to capital in a “take-what-you-want” manner.

数据殖民主义对全球政治活动产生了深远影响。在政治活动中，算法被用于分析选民数据，制定精准的竞选策略和宣传内容，甚至制造和传播政治谣言，影响选民的投票。这种依赖使得政府决策更加依赖科技公司的数据和技术，进一步加强了科技巨头的影响力，从而在全球范围内巩固了其新殖民主义的控制机制。数据殖民主义还对社会结构和经济不平等产生了深远影响。通过数据关系这种新型的人类关系，数据可以被提取出来并进行商品化。全球范围内的社会生活已成为可供其提取的“开放”资源，以某种“予取予求”的方式供资本任意取用。

II. Embedding Mechanisms: Deep Implantation of Western Values into Algorithm Design

二、嵌入机制：西方价值理念深度植入算法设计

Algorithms are not merely a set of cold computational rules; they carry the values and cultural tendencies of their designers. The designers and users of algorithms often unconsciously encode specific cultural preferences, political stances, and economic interests into the design logic of algorithms, thereby promoting and consolidating Western cultural hegemony on a global scale. As algorithms are widely applied across various fields—from content recommendations on social media to news media coverage choices and targeted online advertising—the embedding of Western values has a profound impact on the global public opinion landscape.

算法不仅仅是一系列冷冰冰的计算规则，它们承载着设计者的价值观念和文化倾向。算法的设计者和使用者，往往在不自觉中将特定的文化偏好、政治立场和经济利益等编码进算法的设计逻辑之中，从而在全球范围内推广和巩固西方文化霸权。随着算法在各个领域的广泛应用，从社交媒体的内容推荐到新闻媒体的报道选择，再到在线广告的定向投放，西方价值理念的嵌入对全球舆论格局产生了深远影响。

(1) Creating Algorithmic Bias: Suppressing Diverse Cultural Needs

（一）制造算法偏见：压抑多元文化需求

When designing and applying algorithms, Western countries often prioritize content that aligns with Western aesthetics and lifestyles, while neglecting or marginalizing other cultures, thereby triggering bias against specific individuals or groups. The causes of algorithmic bias are complex and diverse, mainly including the following aspects: First, bias in “datasets.” The training process of algorithms relies on massive amounts of data. If certain groups are underrepresented or have low sample sizes in the training “dataset,” the algorithm may develop bias against these groups. Common types of bias include gender bias, racial bias, regional bias, and occupational bias. Some scholars argue: “Algorithms solve problems based on fixed logic set by humans, which usually can only describe universally applicable empirical facts while ignoring small-probability events in reality. This easily leads to the algorithmic exaggeration of empirical correlations as necessary causal relationships, making algorithmic logic overly absolute and ultimately resulting in algorithmic discrimination against minority groups.” [4]4 See Chen Changfeng and Zhang Xin, “The Generation, Discrimination, and Regulation of Algorithmic Bias,” *Journal of Social Sciences*, no. 1 (2023), p. 12. Second, flaws in the algorithm design itself—the subjective logic of designers may inadvertently implant bias. For example, in 2018, Amazon’s automated recruitment algorithm tended to give lower scores to resumes containing the word “female,” implicitly reflecting the values of the algorithm’s designers and trainers. Third, bias in the application environment. Algorithms may exhibit different effects and impacts in diverse application scenarios, and this environmental sensitivity requires algorithms to be carefully adjusted and optimized in specific contexts.

西方国家在算法设计和应用时，往往优先推荐符合西方审美和生活方式的内容，而忽视或边缘化其他文化，从而引发对特定个体或群体的偏见。算法偏见的成因复杂多样，主要包括以下几个方面：一是“数据集”的偏见。算法的训练过程需要依赖海量数据。如果训练所用的“数据集”中某些群体的代表性不足或样本量偏低，算法可能对这些群体形成偏见。常见的偏见类型包括性别偏见、种族偏见、地域偏见以及职业偏见等。有学者认为：“算法依据人类所设置的固定逻辑解决问题，这种固定逻辑通常只能描述普遍适用的经验事实，而忽略现实中的小概率事件。这很容易导致经验上的相关性被算法夸大为必然存在的因果关系，造成算法逻辑过于绝对，最终导致对少数群体的算法歧视。”二是算法设计本身存在缺陷，设计者的主观逻辑可能在不经意间植入偏见。如2018年亚马逊公司的自动招聘算法倾向于给含“女”字的简历打低分，隐含着算法设计者及训练者的价值观。三是算法应用环境偏差。算法在多样化的应用场景中可能会展现出不同的效果和影响，这种环境敏感性要求算法必须在具体情境中进行细致的调整和优化。

Julia Angwin, a journalist for the Western public media outlet ProPublica, and her team, through in-depth investigation, exposed many hidden cases of algorithmic bias. In an analysis of data on over 7,000 criminal defendants in Florida, USA, they found that an algorithm widely used in the U.S. judicial system exhibited racial bias, misclassifying African Americans as high-risk offenders at a rate 45% higher than other ethnicities, even after accounting for factors such as age, gender, criminal record, and likelihood of future crime. [5]5 Julia Angwin et al., “Machine Bias,” ProPublica, May 23, 2016, https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing. In another study by ProPublica, Julia Angwin posted an ad on Meta targeting “first-time homebuyers” and “people likely to move,” but specifically excluded groups with racial similarity to “African Americans,” “Asian Americans,” and “Hispanics.” Although this practice violated the U.S. Fair Housing Act, Meta still accepted the ad. By excluding users with specific racial similarities, Meta’s advertising system effectively engaged in racial discrimination. Thus, algorithmic bias tends to promote and amplify the values and preferences of mainstream culture while ignoring the uniqueness and diversity of other cultures. In particular, many algorithms and AI systems are developed by Western companies, unconsciously reflecting Western values and preferences, becoming powerful tools for maintaining Western algorithmic hegemony.

西方公共媒体ProPublica的记者朱莉娅·安格温（Julia Angwin）及其团队通过深入调查，揭露了许多隐蔽的算法偏见案例。在对美国佛罗里达州超过7000名刑事被告的数据分析中，他们发现一个广泛用于美国司法系统的算法存在种族偏见，该算法将非裔美国人错误地归类为高风险犯罪者的概率高出其他族裔45%,即便已考虑了年龄、性别、犯罪记录和未来犯罪可能性等因素。在ProPublica进行的另一项研究中，朱莉娅·安格温在Meta上发布了一则针对“首次购房者”和“可能搬家者”的广告，但特别排除了与“非裔美国人”“亚裔美国人”和“拉美裔”有种族相似性的人群。尽管这种做法违反了美国的《公平住房法》,Meta仍旧接受了这则广告。通过排除具有特定种族相似性的用户，Meta的广告系统实际上进行了种族歧视。由此可见，算法偏见倾向于推广和放大主流文化的价值观和偏好，而忽视其他文化的独特性和多样性。特别是许多算法和人工智能系统是由西方国家公司开发，它们不自觉地反映了西方的价值观和偏好，成为维护西方算法霸权的有力武器。

At present, with the mainstreaming of identity politics elements in Western societies and the resurgence of conservatism, the disciplining of algorithms in service of bourgeois rule has taken on more complex characteristics. Although the combination of identity politics and algorithms ostensibly manifests as respect for multiculturalism, its essence remains a strategy by the ruling class to divide the ruled class. By accepting seemingly harmless subcultures, the ruling class is able to maintain its class rule and thereby open up new profit growth points. In this sense, identity politics and conservatism, in that algorithmic discipline serves the bourgeoisie, exhibit a convergence of strategies, jointly safeguarding bourgeois interests and the existing social order.

当前，随着西方社会身份政治元素的主流化与保守主义的回潮，使得算法规训在服务于资产阶级统治的过程中呈现出更为复杂的特点。尽管身份政治与算法的结合表面上体现为对多元文化的尊重，但其实质仍然是统治阶级分化被统治阶级的策略。通过接受那些看似无害的亚文化，统治阶级得以维系其阶级统治，并借此开拓新的利润增长点。在这个意义上，身份政治与保守主义在算法规训服务于资产阶级这一点上，呈现出殊途同归的策略，共同维护着资产阶级的利益和现有的社会秩序。

(II) Dominating Agenda Setting: Precisely Selecting Western Discourse Topics

（二）主导议题设置：精准选取西方话语议题

In Plato’s *Republic*, the allegory of the cave vividly reveals the complex relationship between the pseudo-environment and reality. The imprisoned cave dwellers can only perceive the world through shadows on the wall, mistaking these shadows for reality itself. Yet when a cave dweller is freed and confronts the real world directly, he truly understands the essence of reality. Plato uses this allegory to point out that what we ordinarily take for reality may be merely a distorted reflection of a higher-level reality. Some scholars argue that the development of media technology in the intelligent age is shaping a similar pseudo-environment, “seemingly providing us with a clearer information environment, but in fact further exacerbating the distortion of reality.” [6]6 Zhang Hua, “The Pseudo-Environment in the Intelligent Age and Its Impact on Public Cognition,” *Journal of Communication Studies*, no. 3 (2022): 45. Through selective reporting, Western media carefully select Western discourse topics, shaping the public’s perception and understanding of the world, yet this perception often deviates from objective reality.

在柏拉图的《理想国》中，洞穴比喻生动地揭示了拟态环境与现实之间的复杂关系。被囚禁的洞穴居民只能通过墙壁上的影子来感知世界，他们将这些影子误认为是现实本身。然而当洞穴居民获得解放、直面真实的世界时，他才真正理解了现实的本质。柏拉图用这个比喻指出，日常我们所认为的现实，可能仅仅是更高层次现实的一个扭曲反映。有学者认为，智能时代媒介技术的发展，正塑造了类似的拟态环境，“看似为我们提供了一个更为清晰的信息环境，实际却进一步加剧了对现实的扭曲程度”。西方媒体通过选择性报道，精心选取西方话语议题，塑造了公众对世界的认知和理解，但这种认知往往与客观现实相偏离。

Within the pseudo-environment constructed by Western ideology, the concepts and discourses of Western-centrism are quietly embedded into algorithms. On the one hand, Western states meticulously construct algorithmic narrative rules, setting the dominant topics in international discussion by prioritizing content consistent with Western values and national interests. At the domestic level, to shape and beautify the national image and create the illusion of freedom and democracy, Western states use algorithmic technology to conduct measurement-based monitoring and processing of news searches and news hotspots. At the external level, Western states, through algorithmic recommendation technology, carefully package their own political ideas and international propositions for cross-border dissemination, aiming to promote Western “universal values” and construct and maintain a positive international image. On the other hand, Western states rely on algorithmic monopoly to vigorously slander and denigrate the images of other countries. Through the targeted operation of algorithms, Western states can selectively emphasize or amplify certain negative information in information dissemination, thereby creating an unfavorable public opinion environment for other countries on the international stage. In cultural products, Western states promote cultural products such as films, books, and music that reflect Western values, shaping stereotypical impressions of China and other countries in the international cultural market. For example, on economic and trade issues, Western media use algorithms to emphasize China’s trade surplus and market access issues, as well as the competition they bring to domestic industries, while paying less attention to China’s contributions to global economic growth and its measures to open markets.

在西方意识形态营造的拟态环境下，西方中心主义的理念和话语悄无声息地植入到算法中。一方面，西方国家精心构建算法叙事规则，通过优先展示与西方价值观和国家利益相一致的内容，设定了国际讨论中的主导议题。在对内层面，为了塑造和美化国家形象，营造自由民主假象，西方国家借助算法技术对新闻搜索和新闻热点进行测量性监控和处理。在对外层面，西方国家通过算法推荐技术，将自身政治理念和国际主张精心包装，进行跨国传播，旨在推广西方“普世价值”观，构建和维护其正面国际形象。另一方面，西方国家又依托算法垄断，竭力污蔑和贬损他国形象。通过算法的定向操作，西方国家能够在信息传播中选择性地强调或放大某些负面信息，从而在国际舞台上对他国形成不利的舆论环境。在文化产品上，西方国家推广反映西方价值观的电影、书籍和音乐等文化产品，在国际文化市场中塑造对中国等国的刻板印象。如在经济和贸易议题上，西方媒体通过算法强调中国的贸易顺差和市场准入问题，及其给本国产业带来的竞争，而较少关注中国对全球经济增长的贡献和开放市场的举措。

(III) Utilizing Algorithmic Filtering: Selectively Accepting and Filtering Information

（三）利用算法筛选：选择性接受和过滤信息

The American scholar Kurt Lewin proposed the famous gatekeeping theory, whose core idea is that in the process of information dissemination, there exist key nodes or individuals who control the flow of information, deciding which information the public can access and which information is screened or filtered out. Gatekeeping theory provides us with an important tool for understanding the power structure and distribution of influence in the information dissemination process. As algorithmic recommendation systems increasingly dominate information dissemination, the operational mechanisms and decision-making processes of these technologies have become a new domain for gatekeeping theory. Western states use this new type of gatekeeping mechanism, through algorithmic recommendation systems and social media platforms, to influence the public’s information selection and dissemination.

美国学者库尔特·卢因（Kurt Lewin）提出了著名的把关人理论，其核心观点是，在信息传播的过程中，存在着一些关键的节点或个体，他们控制着信息的流向，决定哪些信息可以被公众获知，哪些信息则被筛选或过滤掉。把关人理论为我们提供了一个理解信息传播过程中权力结构和影响力分布的重要工具。随着算法推荐系统在信息传播中的主导作用不断增强，这些技术的运作机制和决策过程成为把关人理论的新领域。西方国家利用这种新型的把关人机制，通过算法推荐系统和社交媒体平台，影响着公众的信息选择和传播。

Research by the American data scientist Cathy O’Neil has found that Meta uses complex algorithms to determine the content users see in their news feeds, including petitions, news reports, personal updates, and more. O’Neil argues: “Facebook looks like a modern town square, but it can decide what we see and learn on its platform based on its own interests.” [7]7 Cathy O’Neil, *Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy* (New York: Crown, 2016), 145. This kind of algorithm not only shapes our information horizon but also has the capacity to guide public opinion. For example, by prioritizing the display of petitions aligned with specific political agendas, it influences the direction of political topics. Meta also uses advanced language recognition technology to classify information as positive or negative, then conducts group experiments on users: reducing negative content in the information flow for half the users, while reducing positive content for the other half. In this way, Meta demonstrates its enormous influence. Yet it is not the only company with such capabilities; other tech giants, including Google, Apple, Microsoft, Amazon, and major mobile service providers such as Verizon and AT&T, all possess vast amounts of user behavioral data and have multiple means to influence public choices and decisions.

美国数据科学家凯西·奥尼尔（Cathy O'Neil）的研究发现，Meta使用复杂的算法来决定用户在其新闻推送中看到的内容，包括请愿书、新闻报道、个人动态等。奥尼尔认为：“脸书看起来就像是一个现代化的城镇广场，但它可以根据自身的利益决定我们在其平台上看到和学习到的内容。”这种算法不仅塑造了我们的信息视野，还有能力引导公众舆论。例如，通过优先展示符合特定政治议程的请愿书，影响政治话题走向。Meta还利用先进的语言识别技术，对信息进行正负面分类，然后对用户进行分组实验：对一半用户减少信息流中的负面内容，而对另一半用户则减少正面内容。通过这种方式，Meta展示了其巨大的影响力。然而它并非唯一拥有这种能力的公司，其他科技巨头，包括谷歌、苹果、微软、亚马逊，以及主要的手机服务供应商如威瑞森和美国电话电报公司等，都掌握着海量的用户行为数据，并拥有多种手段来影响公众的选择和决策。

In summary, Western states, by using algorithms to control the classification and distribution of information, allow information that aligns with their interests to flow unimpeded, while restricting information that contradicts them, thereby shaping a public opinion environment favorable to their own values and policy goals in global information dissemination. Under the capitalist system, algorithms become executors of the logic of capital, with their design principles oriented toward capital accumulation and efficiency enhancement. For example, news recommendation algorithms tend to favor low-quality or sensational content that quickly captures user attention over more in-depth and valuable reporting. Similarly, in coverage related to China, Western media selectively highlight negative news, such as environmental pollution and human rights issues, while reporting less on China’s achievements in poverty reduction, technological innovation, and infrastructure construction. Over time, Western audiences gradually lose the ability for deep understanding and critical thinking on complex issues, thereby affecting the intellectual level and quality of rational discussion in the entire society.

总之，西方国家借助算法对信息分发进行分类把控，符合其利益需求的信息畅通无阻，而与之相悖的信息则受到限制，从而在全球信息传播中塑造了一种有利于自身价值观和政策目标的舆论环境。在资本主义体系下，算法成为资本逻辑的执行者，其设计原则以资本积累和效率提升为导向。例如，新闻推荐算法更倾向于推荐那些能迅速吸引用户注意的低质量或煽动性内容，而非更有深度和价值的报道。又如在涉及中国的报道中，西方媒体会选择性地突出报道负面新闻，如环境污染、人权问题等，而较少报道中国在减贫、科技创新、基础设施建设等方面的成就。长此以往，西方受众会逐渐失去对复杂问题的深入理解和批判性思考的能力，从而影响到整个社会的智识水平和理性讨论的质量。

III. Monopoly Mechanism: Building a “Technology-Data-Capital” Integrated Algorithmic Hegemony System

三、垄断机制：打造“技术-数据-资本”多元一体的算法霸权体系

With the rapid development of digital technology, we are witnessing the rise of a new type of algorithmic monopoly mechanism—a hegemonic system centered on algorithms, integrating “technology, data, and capital.” This system consists of several key elements: the research, development, and application of cutting-edge technology; the collection and analysis of massive data; and the investment and operation of substantial capital. These elements interact and reinforce each other, forming a self-reinforcing cycle that enables enterprises possessing these resources to occupy an unassailable position in the global market.

随着数字技术的飞速发展，我们见证了一种新型算法垄断机制的崛起——一个以算法为核心的“技术-数据-资本”多元一体的霸权体系。这一体系由几个关键要素构成：尖端技术的研发与应用，海量数据的收集与分析，以及雄厚资本的投入与运作。这些要素相互作用、相互加强，形成了一个自我强化的循环，使得拥有这些资源的企业能够在全球市场中占据不可撼动的地位。

(I) Technological Monopoly: Using Algorithmic Barriers to Consolidate Hegemonic Status

（一）技术垄断：利用算法壁垒巩固霸权地位

The Egyptian scholar Samir Amin argues that the capitalist world is a system composed of a “center” and a “periphery,” two “socio-economic formations.” [8]8 Samir Amin, *Accumulation on a World Scale: A Critique of the Theory of Underdevelopment* (New York: Monthly Review Press, 1974), 15. In this system, developed capitalist countries, by virtue of their first-mover advantages in technology, capital, and market access, occupy the central position in the world economic system, forming “the economic domination of the center over the periphery.” [9]9 Ibid., 23. Late-developing countries, in the process of catching up and developing, often find themselves trapped in a state of dependency on developed countries. This dependency is reflected not only in traditional economic fields but also in the emerging digital economy. Considering the complexity of algorithmic technology development and the long-term advantages Western countries have cultivated in the algorithmic field, late-developing countries face the challenge of algorithmic dependency. This dependency limits their ability to achieve autonomy and innovation in the rapidly developing digital economy, deepening their disadvantageous position in global technological competition.

埃及学者萨米尔·阿明（Samir Amin）认为，资本主义世界是由一个“中心”和“外围”两种“社会经济形态”构成的体系。在这个体系中，发达资本主义国家凭借其在技术、资本和市场准入上的先发优势，占据了世界经济体系的中心地位，形成了“中心对外围的经济统治”。后发国家在追赶发展的过程中，往往发现自己陷入了对发达国家的依附状态。这种依附不仅体现在传统经济领域，也体现在新兴的数字经济领域。考虑到算法技术发展的复杂性以及西方国家在算法领域的长期耕耘优势，后发国家面临着算法依赖的挑战。这种依赖限制了它们在快速发展的数字经济中实现自主和创新的能力，加深了它们在全球技术竞争中的不利地位。

In terms of the mathematical foundation, the core technology of algorithms, Western countries have already accumulated a significant advantage. Western countries recognized and established their own algorithmic advantage from a mathematical foundation relatively early. As British researcher Marcus du Sautoy pointed out, algorithms “are, in a sense, mathematical operations.” [10]10 Marcus du Sautoy, *The Creativity Code: Art and Innovation in the Age of AI* (Cambridge, MA: Harvard University Press, 2019), p. 6. Mathematics provides a precise language and tool, enabling algorithms to process information in a rigorous and systematic manner. Alan Turing, the father of artificial intelligence, believed that an intelligent machine is essentially “a mathematical ideal form, rather than an actual machine.” [11]11 Alan Turing, “Computing Machinery and Intelligence,” *Mind* 59, no. 236 (1950): 433–460. Walter Pitts and Warren McCulloch demonstrated through their research that even a simple model simulating the neurons of the human brain could “in principle compute a certain type of logical function.” [12]12 Warren S. McCulloch and Walter Pitts, “A Logical Calculus of the Ideas Immanent in Nervous Activity,” *Bulletin of Mathematical Biophysics* 5, no. 4 (1943): 115–133. Thus, it can be seen that the design and implementation of intelligent algorithms, whether at the software or hardware level, are fundamentally inseparable from the support of mathematical models and logical reasoning. The deep accumulation of Western countries in this area provides a solid foundation for the development of their algorithmic technology.

在算法的核心技术——数学基础方面，西方国家已经积累了显著优势。西方国家较早地从数学基础方面认识和确立了自身的算法优势地位。正如英国研究者马库斯·杜·索托伊（Marcus du Sautoy）所指出的，算法“在某种程度上，它们就是数学运算”。数学提供了一种精确的语言和工具，使得算法能够以一种严格和系统化的方式处理信息。人工智能之父艾伦·图灵（Alan Turing）认为，智能机器本质“是一种数学上的理想形式，而不是一台实际的机器”。沃尔特·皮茨（Walter Pitts）和沃伦·麦卡洛克（Warren McCulloch）通过他们的研究证明，即使是模拟人脑神经元的简单模型，“原则上能够计算一定类型的逻辑函数”。由此可见，智能算法的设计和实现，无论是在软件还是硬件层面，其核心都离不开数学模型与逻辑推理的支撑。西方国家在这方面的深厚积累，为其算法技术的发展提供了坚实的基础。

Another mechanism of algorithms—the algorithmic black box mechanism—is also strengthening the algorithmic hegemonic position of Western countries. Frank Pasquale, a law professor at the University of Maryland, believes that with the proliferation of online platforms such as social media, the transparency of personal life is increasing, while at the same time, the algorithmic tools developed by Wall Street and Silicon Valley companies are little known. He points out: “This phenomenon reveals the paradox of the information age: although the breadth and depth of data are increasingly astonishing, its ‘black box’ nature makes it difficult for us to see through the logic behind it, and some information that can have an important impact on us is only accessible to insiders.” [13]13 Frank Pasquale, *The Black Box Society: The Secret Algorithms That Control Money and Information* (Cambridge, MA: Harvard University Press, 2015), p. 2. Due to information asymmetry and the opacity of algorithmic processes, public ignorance of algorithmic decision-making processes is exacerbated, making it difficult for the public to understand the logic and potential biases behind algorithms. In Western countries, the “black box” nature of algorithms has become a covert hegemonic tool of political propaganda, quietly shaping voters’ opinions outside the public eye. During the 2016 U.S. presidential election, the strategy of the data analysis company Cambridge Analytica was a notable case of this phenomenon. The company claimed that its data-driven approach played a key role in Trump’s victory. By applying the Big Five personality model, Cambridge Analytica claimed to be able to conduct in-depth behavioral analysis of millions of American voters, creating fine-grained personality profiles that went beyond traditional demographics. This incident highlighted the powerful influence of algorithms in modern society, while also revealing the risks they may pose in the absence of transparency and regulation.

算法的另一机制——算法黑箱机制也在强化西方国家的算法霸权地位。美国马里兰大学法学教授弗兰克·帕斯奎尔（Frank Pasquale）认为，随着社交媒体等在线平台的普及，个人生活的透明度日益增加，而与此同时，由华尔街和硅谷公司开发的算法工具却鲜为人知。他指出：“这种现象展现了信息时代的悖论：尽管数据的广度和深度越来越惊人，但其‘黑箱’特性却让我们难以洞察其背后的逻辑，有些会对我们产生重要影响的信息却只有局内人才能获得。”由于信息不对称和算法过程的不透明性，加剧了公众对算法决策过程的无知，使得公众难以理解算法背后的逻辑和潜在偏见。在西方国家，算法的“黑箱”特性已成为政治宣传的一种隐性霸权工具，它在公众视线之外悄然塑造着选民的意见。2016年美国总统选举期间，数据分析公司剑桥分析（Cambridge Analytica）的策略便是这一现象的显著案例。该公司声称，其基于数据的方法对特朗普的胜选起到了关键作用。通过应用五大人格模型，剑桥分析声称能够对数百万美国选民进行深入的行为分析，创建出超越传统人口统计学的精细人格画像。这一事件凸显了算法在现代社会中的强大影响力，同时也揭示了其在缺乏透明度和监管的情况下可能带来的风险。

(2) Data Monopoly: Controlling the Flow of Global Data Resources

（二）数据垄断：控制全球数据资源流动

Yuval Noah Harari, in his works, proposed the concept of “dataism,” arguing that in an intelligent society, the centrality and importance of data are increasingly prominent, almost forming a new “religion.” Harari emphasizes: “Dataism began as a neutral scientific theory, but is gradually becoming a religion that judges right and wrong.” [14]14 Yuval Noah Harari, *Homo Deus: A Brief History of Tomorrow* (New York: Harper, 2017), p. 351. In this system, data and algorithms may replace traditional deities and humans, becoming objects of worship and dependence. Within the framework of dataism, data itself is endowed with a supreme status, becoming the key to understanding the world. Western countries, relying on their deep accumulation in technology and algorithms, have already taken a leading position in the wave of dataism. Through their advanced data processing capabilities and algorithmic advantages, they have not only achieved great success in the economic sphere but also played a significant role in shaping global information flows and cultural influence.

尤瓦尔·赫拉利（Yuval Noah Harari）在其著作中提出了“数据主义”这一概念，认为在智能社会中，数据的中心地位和重要性日益凸显，几乎形成了一种新的“宗教”。赫拉利强调：“数据主义一开始也是一个中立的科学理论，但正逐渐成为要判别是非的宗教。”在这种体系中，数据和算法可能会取代传统的神祇和人类，成为人们崇拜和依赖的对象。在数据主义的框架下，数据本身被赋予了至高无上的地位，成为理解世界的关键。西方国家凭借其在技术和算法领域的深厚积累，已经在数据主义的浪潮中占据了领先地位。它们通过先进的数据处理能力和算法优势，不仅在经济领域取得了巨大成功，也在塑造全球信息流动和文化影响力方面发挥了重要作用。

Large Western technology companies have accumulated vast amounts of user data through their platforms and business models, forming a trend of data centralization. In the era of the digital economy, data has become a key factor of production, and the ability to control data is directly related to a company’s competitiveness and market position. Specifically, Western countries control global data resources mainly through the following methods: First, building super data centers. The United States possesses the world’s largest data centers; the data center built by the U.S. National Security Agency can store 100 years of global electronic communication data. [15]15 James Bamford, *The Shadow Factory: The Ultra-Secret NSA from 9/11 to the Eavesdropping on America* (New York: Doubleday, 2008), p. 320. Second, conducting global-scale data surveillance. The United States, leveraging the internet that originated within its borders, and with advantages in technology, standards, and domain name resolution services, has established a massive data surveillance system. Western media have revealed that the U.S. National Security Agency remotely stole 97 billion emails and 124 billion telephone records globally within 30 days, including 500 million from Germany, 70 million from France, and 60 million from Spain. [16]16 “NSA Prism program taps in to user data of Apple, Google and others,” *The Guardian*, June 7, 2013. Third, using technological monopolies to create tools for data collection. Due to the United States’ first-mover advantage in chip design, operating systems, and positioning systems, the U.S. has meticulously built a global data collection network. Using the GPS positioning system, the U.S. can precisely grasp the location information of people and objects. With the proliferation of smart cars, vehicle driving systems can collect vast amounts of information data through cameras and lidar.

西方大型科技公司通过其平台和业务模式积累了大量的用户数据，形成了数据集中化的趋势。在数字经济时代，数据已成为一种关键的生产要素，对数据的控制能力直接关系到企业的竞争力和市场地位。具体而言，西方国家主要通过以下方式控制全球数据资源：一是打造超级数据中心。美国拥有世界上最大的数据中心，美国国家安全局建造的数据中心，可存储100年的全球电子通信数据。二是进行全球范围的数据监控。美国利用发源于其国内的互联网，凭借技术、标准和域名解析服务等优势，建立庞大的数据监控体系。西方媒体透露，美国国家安全局曾在30天内远程窃取全球970亿封邮件和1240亿条电话数据，其中德国5亿份、法国7000万份、西班牙6000万份。三是利用技术垄断打造数据收集利器。由于美国在芯片设计、操作系统、定位系统的先发优势，美国精心打造覆盖全球的数据收集网络。利用GPS定位系统，美国可精准掌握人员和物体的位置信息。随着智能汽车的普及，通过摄像头和激光雷达，车辆驾驶系统可收集到海量的信息数据。

(3) Capital Monopoly: Forming a Global Digital Monopoly Alliance

（三）资本垄断：形成全球数字垄断同盟

In the wave of the digital age, the global flow and concentration of capital have reached an unprecedented scale, giving rise to a new type of global influence—the global digital monopoly alliance. This phenomenon is a key transformation undergone by the capitalist economy driven by digital technology, in which a few tech giants, leveraging their abundant capital, cutting-edge technology, and extensive market influence, have established a powerful digital monopoly position on a global scale. Through carefully orchestrated mergers and acquisitions, patent strategies, and market operations, these enterprises continuously consolidate and expand their market power, ultimately forming a cross-border digital monopoly alliance. They not only dominate the economic sphere but also exert a profound influence on multiple levels, including culture, politics, and society. Emerging digital media and social media platforms are aggressively expanding their territory in the global digital domain, posing severe challenges to the information security of developing countries.

在数字化时代的浪潮中，资本的全球流动和集聚达到了空前的规模，催生了一种新型的全球性影响力——全球数字垄断同盟。这一现象是资本主义经济在数字技术推动下经历的关键转变，其中少数科技巨头利用其雄厚的资本、尖端的技术和广泛的市场影响力，在全球范围内确立了强大的数字垄断地位。这些企业通过精心策划的并购活动、专利策略和市场运作，不断巩固并扩展其市场势力，最终形成了一个跨越国界的数字垄断联盟。它们不仅在经济领域占据主导地位，更在文化、政治和社会等多个层面产生了深远的影响。新兴的数字媒体和社交媒体平台在全球数字领域攻城略地，对发展中国家的信息安全带来了严峻挑战。

Against the backdrop of capital monopoly, enterprises within the global digital monopoly alliance often use algorithms to collude, reaching agreements on price fixing and the like. Algorithmic hegemony brings enormous economic benefits to Western countries. Enterprises that master advanced algorithms can optimize production processes and improve operational efficiency through algorithms, thereby gaining an advantage in market competition. High-frequency trading algorithms in the financial sector can use their rapid data processing and decision-making capabilities to obtain high profits in the stock market. Additionally, algorithms are used for precision advertising, enabling advertisers to deliver more targeted ads based on users’ interests and behavioral data, thereby improving advertising effectiveness and revenue. Enterprises exchange information and coordinate prices through algorithms, causing market prices to be artificially manipulated at supra-competitive levels. This algorithmic collusion not only restricts market competition but also excludes other competitors from entering the market, leading to an unfair distribution of market resources. When consumers face market prices manipulated by a monopoly alliance, they cannot make optimal consumption decisions based on real market supply and demand, and their consumer rights are severely damaged. The global digital monopoly alliance formed by capital monopoly also suppresses market innovation vitality through algorithmic discipline. Monopoly enterprises use their algorithmic advantages to suppress and exclude emerging enterprises and innovative products. When an emerging enterprise launches an innovative product similar to or competitive with that of a monopoly enterprise, the monopoly enterprise may adjust search rankings, recommendation weights, and other means through algorithms to reduce the exposure and market opportunities of the emerging product. This behavior not only hinders fair market competition but also further solidifies the market position of the monopoly enterprise. Unlike the direct control of territorial presence by traditional state power, data control and network influence have become new symbols of power.

在资本垄断的背景下，全球数字垄断同盟中的企业往往利用算法进行共谋，达成价格垄断等协议。算法霸权为西方国家带来了巨大的经济利益。掌握先进算法的企业能够通过算法优化生产流程、提高运营效率，从而在市场竞争中占据优势。金融领域的高频交易算法能够利用其快速的数据处理和决策能力，在股票市场中获取高额利润。此外，算法还被用于精准广告投放，使广告商能够根据用户的兴趣和行为数据，投放更具针对性的广告，提高广告效果和收益。企业之间通过算法交换信息、协调价格，使得市场价格被人为操控在超竞争水平。这种算法共谋不仅限制了市场竞争，还排除了其他竞争者进入市场的机会，导致市场资源的不公平分配。消费者在面对垄断同盟操控的市场价格时，无法根据真实的市场供需关系做出最优的消费决策，其消费权益受到严重损害。资本垄断形成的全球数字垄断同盟，通过算法规训还抑制了市场的创新活力。垄断企业利用其算法优势，对新兴企业和创新产品进行打压和排挤。当新兴企业推出与垄断企业产品相似或具有竞争力的创新产品时，垄断企业可能会通过算法调整搜索排名、推荐权重等方式，降低新兴产品的曝光度和市场机会。这种行为不仅阻碍了市场的公平竞争，还使得垄断企业的市场地位得以进一步固化。与传统国家权力对领土在场的直接控制不同，数据控制和网络影响力成为新的权力象征。

IV. Algorithmic Sovereignty and Cultural Autonomy: Strategic Pathways to Break Western Algorithmic Hegemony

四、算法主权与文化自主：打破西方算法霸权的策略路径

The era of artificial intelligence is an era full of challenges, but also an era full of opportunities. The algorithmic hegemony established by the West by virtue of its first-mover advantage is quietly influencing the cognition and decision-making of billions of people around the world in cultural and ideological terms. The information cocoons, cultural homogenization, and impacts on local industries brought about by algorithmic hegemony compel us to re-examine and seek strategic pathways to break through this hegemony. To break hegemony, national autonomy and capacity building are crucial. Developed countries have more transnational enterprises in the field of artificial intelligence, and the AI governance platforms they dominate started earlier and have greater influence. In contrast, many developing countries lack leading AI technology enterprises, and their voice, representation, and discursive power are weaker in the global governance process. Therefore, the state needs to increase investment in the field of artificial intelligence, encourage cooperation among universities, research institutions, and enterprises, overcome key core technical difficulties, and cultivate a group of locally based AI enterprises with international competitiveness, thereby reducing dependence on Western algorithms and enhancing autonomous discursive power in the international algorithmic field.

人工智能时代是一个充满挑战的时代，也是一个充满机遇的时代。西方凭借其先发优势建立的算法霸权，在文化和意识形态上悄然影响着全球数十亿人的认知和决策。算法霸权带来的信息茧房、文化同质化以及对本土产业的冲击，促使我们必须重新审视并寻求突破这一霸权的策略路径。为打破霸权，国家自主性与能力建设至关重要。发达国家在人工智能领域拥有较多跨国企业，主导的人工智能治理平台起步较早、影响力较大。相较之下，许多发展中国家缺少领先的人工智能科技企业，在全球治理进程中声势较弱、代表性和话语权不足。因此，国家需加大对人工智能领域的投入，鼓励高校、科研机构与企业开展合作，攻克关键核心技术难题，培育一批具有国际竞争力的本土人工智能企业，从而减少对西方算法的依赖，增强在国际算法领域的自主话语权。

(I) Increase investment in scientific and technological research and development, and overcome core algorithmic technologies

（一）加大科技研发投入，攻克算法核心技术

Fourth, Algorithmic Sovereignty and Cultural Autonomy: Strategic Pathways to Break Western Algorithmic Hegemony

第一，设立专项研发基金。国家应设立专门的人工智能研发基金，用于支持高校、科研机构和企业在算法领域的基础研究和前沿技术探索。启动“核心算法专项计划”,集中力量攻克深度学习、强化学习、迁移学习等关键算法的理论难题和应用瓶颈，为我国算法发展提供坚实的技术基础。第二，推动产学研深度融合。鼓励高校、科研机构与企业建立紧密的合作关系，共建联合实验室、研究中心等创新平台。通过产学研合作，加速算法研究成果的转化应用，推动算法技术在实际场景中的落地。比如，企业可以为高校和科研机构提供实际数据和应用场景，而高校和科研机构则为企业解决技术难题，共同开发具有自主知识产权的先进算法，实现资源共享、优势互补。

The era of artificial intelligence is both an era of challenges and an era of opportunities. The algorithmic hegemony established by the West through its first-mover advantage is quietly influencing the cognition and decision-making of billions of people worldwide in cultural and ideological dimensions. The information cocoons, cultural homogenization, and impacts on local industries brought about by algorithmic hegemony compel us to re-examine and seek strategic pathways to break this hegemony. To break this hegemony, national autonomy and capacity-building are crucial. Developed countries possess more multinational enterprises in the field of artificial intelligence and have dominated AI governance platforms that started earlier and have greater influence. In contrast, many developing countries lack leading AI technology enterprises, have a weaker voice in global governance processes, and suffer from insufficient representation and discursive power. Therefore, states need to increase investment in the field of artificial intelligence, encourage cooperation among universities, research institutions, and enterprises, overcome key core technical difficulties, cultivate a batch of locally competitive AI enterprises, thereby reducing dependence on Western algorithms and enhancing autonomous discursive power in the international algorithm field.

（二）完善算法治理与监管体系，保障算法应用的公平、公正与透明

(1) Increase Investment in Scientific and Technological R&D, Overcome Core Algorithm Technologies

习近平总书记指出：“要依法加强网络社会管理，加强网络新技术新应用的管理，确保互联网可管可控，使我们的网络空间清朗起来。”当前，算法的专业性、复杂性和不透明性使得其掌握在少数人手中，导致算法霸权的出现。我国现有监管制度存在缺乏统一的监管部门、监管乏力、法律法规不健全等问题。为此，国家应加快人工智能领域立法进程，明确算法应用的边界和责任主体，建立健全算法备案、评估和问责机制，加强对算法的全流程监管，防止算法滥用对社会公共利益的损害。同时，提升监管能力，利用技术手段对算法进行实时监测和评估，确保算法应用符合国家法律法规和伦理道德标准。

First, establish special R&D funds. The state should set up dedicated AI R&D funds to support basic research and cutting-edge technology exploration in the algorithm field by universities, research institutions, and enterprises. Launch a “Core Algorithm Special Plan,” concentrating efforts to overcome theoretical difficulties and application bottlenecks in key algorithms such as deep learning, reinforcement learning, and transfer learning, providing a solid technical foundation for China’s algorithm development. Second, promote deep integration of industry, academia, and research. Encourage universities, research institutions, and enterprises to establish close cooperative relationships, jointly building innovation platforms such as joint laboratories and research centers. Through industry-academia-research cooperation, accelerate the transformation and application of algorithm research results, promoting the implementation of algorithm technologies in practical scenarios. For example, enterprises can provide universities and research institutions with real data and application scenarios, while universities and research institutions solve technical problems for enterprises, jointly developing advanced algorithms with independent intellectual property rights, achieving resource sharing and complementary advantages.

（三）强化算法科技人才的培养与引进

(2) Improve the Algorithm Governance and Regulatory System, Ensure Fairness, Justice, and Transparency in Algorithm Applications

第一，优化人才培养体系。在高校增设人工智能相关专业和课程，培养一批既懂技术又懂伦理、法律的复合型算法人才。开展人工智能领域的联合培养项目，选派优秀学生到国外顶尖高校深造，引进国外先进的教学理念和课程体系，提高我国算法人才的培养质量。第二，吸引海外人才回国发展。制定优惠政策，吸引海外优秀算法人才回国工作和创业。为海外人才提供优厚的待遇、良好的科研条件和广阔的发展空间，使其能够充分发挥自身优势，为国家算法事业注入新的活力。通过人才的引进，快速提升我国算法领域的研发实力和创新能力。

General Secretary Xi Jinping pointed out: “We must strengthen online social management according to law, strengthen the management of new network technologies and new applications, ensure that the internet is manageable and controllable, and make our cyberspace clear and bright.” [17]17 "Information cocoons" (信息茧房) refers to the phenomenon where algorithms, by filtering content based on user preferences, confine individuals to a narrow information environment, limiting exposure to diverse viewpoints. This term is widely used in Chinese academic and policy discourse on digital governance. Currently, the professionalism, complexity, and opacity of algorithms place them in the hands of a few, leading to the emergence of algorithmic hegemony. China’s existing regulatory system suffers from problems such as

（四）推动建立公平合理的算法国际规则

First, strengthen international cooperation in the algorithmic field. Conduct extensive cooperation with international organizations and other countries in the algorithmic field, and jointly formulate international algorithmic standards and norms. By participating in international AI governance cooperation, enhance the country's discursive power and influence in the international algorithmic field, and promote the international community to jointly address the challenges posed by algorithmic hegemony. Second, actively voice opinions on the international stage, propose algorithmic governance solutions that align with the interests of developing countries, and promote the establishment of a fair, just, open, and inclusive international algorithmic rule system. Advocate for the establishment of an international algorithmic transparency and fairness assessment mechanism, promote the international community to jointly supervise the application of algorithms, prevent the abuse of algorithmic hegemony, and safeguard the fairness and justice of the global algorithmic order.

第一，加强算法领域的国际合作。与国际组织和其他国家在算法领域开展广泛的合作，共同制定国际算法标准和规范。通过参与国际人工智能治理合作，提升国家在国际算法领域的话语权和影响力，推动国际社会共同应对算法霸权带来的挑战。第二，在国际舞台上积极发声，提出符合发展中国家利益的算法治理方案，推动建立公平、公正、开放、包容的国际算法规则体系。倡导建立国际算法透明度和公平性评估机制，促进国际社会共同监督算法的应用，防止算法霸权的滥用，维护全球算法秩序的公平正义。

Online editor: Zhang Jian

网络编辑：张剑