← نسخه فارسی

From Knowledge Production and the Digital Economy to Surveillance, Repression, and Militarization

Artificial intelligence cannot be understood merely as a collection of algorithms, powerful computers, and language models. Just as the Industrial Revolution cannot be explained solely by the invention of the steam engine, the AI revolution cannot be understood without examining the social and political relations within which this technology develops. Artificial intelligence can provide society with unprecedented capabilities in medicine, education, scientific research, agriculture, industry, energy management, transportation, and the rational organization of resources. Yet the same computational capacity can also be used for facial recognition, mass surveillance of citizens, analysis of social networks, control of communications, cyberwarfare, weapons guidance, and the expansion of the capabilities of security institutions. The fundamental question concerning artificial intelligence, therefore, is not simply how advanced a country has become in this technology. The more important questions are: Who controls the technology? Which institutions decide how it is used? Who owns the data and computational infrastructure? And ultimately, which social forces benefit from the value and power that this technology creates?

Examining the development of artificial intelligence in Iran from this perspective is particularly important. In recent years, the Islamic Republic has presented the development of AI as one of the country’s strategic objectives, spoken of creating national AI infrastructure, and attempted to portray progress in this field as part of Iran’s “scientific power” and “technological independence.” But technology does not develop in a political vacuum. When artificial intelligence develops within a structure in which political power is highly concentrated, the free circulation of information is restricted, the internet can be blocked or shut down during periods of crisis, independent political and civic activity faces severe limitations, and security and military institutions exercise extensive power, the direction of AI development cannot be examined independently of the nature of that structure. Under such conditions, the central question is no longer simply whether Iran is “advanced or backward in artificial intelligence.” The question becomes how much of society’s scientific and technological capacity is placed at the service of society itself, and how much of it becomes concentrated within the structure of power.

From the University to Artificial Intelligence: A Society with Significant Human Capital

The roots of artificial intelligence in Iran must first be sought in its universities. Beginning in the 1980s and 1990s, alongside the expansion of computer engineering, electrical engineering, control engineering, and computer science, subjects such as expert systems, pattern recognition, image processing, neural networks, robotics, and intelligent control gradually entered Iranian universities. At that time, there was not yet anything that could reasonably be called an “Iranian artificial intelligence industry,” and activity in the field remained predominantly academic and research-oriented. With the expansion of graduate education in the 2000s, the number of theses, academic papers, and projects related to machine learning, natural language processing, data mining, and computer vision increased, while universities such as the University of Tehran, Sharif University of Technology, Amirkabir University of Technology, Iran University of Science and Technology, and several other academic centers became important hubs of this research.

The idea that Iranian society fundamentally lacks the scientific capacity to enter the age of artificial intelligence is therefore inconsistent with reality. Over several decades, Iran has trained large numbers of engineers, computer specialists, mathematicians, machine-learning researchers, and data-processing experts. A significant portion of this human capital, both inside Iran and after migration, has demonstrated the ability to operate at an international level. A 2025 study published in Iranian Studies by Cambridge University Press similarly emphasizes the contradiction between official narratives of “technological power” and the material realities of technological development. The study shows how projects such as the Simorgh supercomputer, national data centers, and AI-related initiatives have been presented by the government as symbols of technological independence, even though the actual development of AI remains dependent on hardware, capital, supply chains, and the global network of knowledge. This dependence demonstrates that possessing skilled human capital, while essential, is not sufficient to transform a country into an artificial intelligence power.

Iran’s fundamental problem begins precisely here. A society capable of producing scientific talent is not necessarily capable of retaining and fully utilizing that talent. The migration of specialists, restrictions on scientific interaction with the world, difficulties in obtaining advanced computational equipment, sanctions, weak long-term investment, internet restrictions, and an unstable economic environment have collectively created barriers between the human capacity that exists and its transformation into a powerful technological ecosystem. In other words, one of the central contradictions of AI development in Iran is that society has the capacity to produce knowledge and expertise, while the existing economic and political structure restricts the full transformation of that capacity into a productive social force.

The Digital Economy and the Entry of Artificial Intelligence into Everyday Life

The next important stage began in the 2010s with the expansion of the internet, smartphones, and digital platforms. Iranian internet companies, online retailers, ride-hailing services, banks, and financial institutions gradually accumulated enormous volumes of user data, while the use of search algorithms, recommendation systems, customer-behavior analysis, fraud detection, routing, targeted advertising, and data processing expanded. At this stage, artificial intelligence moved beyond the university laboratory and became part of everyday economic activity. This transformation has considerable theoretical importance because it shows that Iran, even without corporations on the scale of Google, Microsoft, or OpenAI, has entered the same global process in which the daily activities of millions of people continuously generate data, and that data, once collected and processed, can become a source of economic value and power.

From this perspective, the user of a platform is not merely a consumer of a service; the user is simultaneously producing data. Every search, purchase, travel route, product selection, amount of time spent on a page, and many other forms of behavior can be converted into data. Algorithms analyze this information and use the results to improve services, predict behavior, target advertising, determine prices, or make business decisions. This transformation represents one of the clearest examples of the socialization of the production process in the digital economy: millions of people, often without necessarily being conscious of their role in this process, participate in producing the raw material of the data economy, while ownership of the infrastructure, algorithms, and economic value resulting from this collective participation remains concentrated in the hands of a limited number of institutions.

The Emergence of Generative AI and the Exposure of Technological Dependence

The emergence of ChatGPT and the wave of generative artificial intelligence beginning in 2022 marked another stage. For the first time, AI use in Iran moved beyond specialized environments, allowing students, translators, writers, physicians, programmers, designers, researchers, and small businesses to work directly with generative models. Yet this widespread adoption also exposed a fundamental reality: most of the advanced artificial intelligence used by Iranian citizens was not developed in Iran. The foundation models, architectures, advanced chips, cloud infrastructure, and enormous capital required to train frontier models are largely concentrated outside the country. We must therefore distinguish among three different levels: Iran possesses significant research capacity and human capital; the application of AI throughout its society and economy is expanding; but in terms of ownership of foundational technologies, advanced computational infrastructure, and the development of frontier models, it remains far behind the leading powers in the field.

This gap has deepened as the AI industry itself has evolved. Modern AI has become increasingly capital-intensive, energy-intensive, and infrastructure-intensive. Training very large models requires data centers, thousands of advanced processors, high-speed networks, enormous quantities of electricity, vast amounts of capital, and continuous interaction with the global ecosystem of science and technology. Under such circumstances, Iran’s restricted access to advanced chips and global capital markets is not merely a commercial problem; it directly affects the country’s ability to develop technology at the frontier. Iran’s situation can therefore be described simply but meaningfully: Iran does not suffer from a shortage of brains; it suffers from structural constraints on transforming brains into productive power.

The Entry of the State and the Transformation of AI into a Question of Power

As the strategic importance of AI became increasingly evident, the Islamic Republic sought to incorporate the technology into official national planning. The adoption of a National Artificial Intelligence Strategy in 2024 and the establishment of institutions intended to direct this field demonstrated that the government no longer viewed AI merely as an academic or economic subject, but as an element of national power. In subsequent years, officials also spoke of developing a national AI platform and infrastructure capable of continuing to operate even under conditions in which Iran’s connection to the global internet might be restricted.

At first glance, a country’s attempt to reduce technological dependence is not necessarily problematic. Developing Persian-language models, keeping sensitive data within the country, and establishing independent computational capacity can even be legitimate and defensible objectives. The problem begins when such a project is placed within the specific political structure of the Islamic Republic—a structure with a long history of internet filtering, restricting the circulation of information, controlling the media, and using technology to monitor society.

It is here that the concept of “national artificial intelligence” acquires a dual meaning. In a democratic structure, national AI infrastructure could serve universities, medicine, education, research, and public services while simultaneously being subjected to public oversight and strict privacy protections. In an authoritarian structure, however, the same infrastructure can further concentrate data and computational power in the hands of the state. Freedom House’s Freedom on the Net 2025 classifies Iran as “Not Free” in terms of internet freedom and points to extensive censorship, digital surveillance, restrictions on access to information, and efforts to direct users toward a domestic network. Under such circumstances, the development of artificial intelligence cannot be separated from the questions of internet freedom and control over information.

The Contradiction Between the Age of Consciousness and the Control of Consciousness

One of the deepest contradictions in the Islamic Republic’s technology policy becomes visible precisely at this point. Artificial intelligence feeds on knowledge, and knowledge requires the circulation of information, freedom to question, criticism, scientific cooperation, access to resources, and unrestricted communication among researchers in order to grow. By contrast, a political system that seeks its survival partly through controlling information, restricting the media, and constraining the independent organization of society comes into conflict with the fundamental logic of a knowledge-based society.

A government cannot, on the one hand, aspire to become an artificial intelligence power while, on the other, separating universities, researchers, and citizens from the free global flow of information. This contradiction does not merely produce inefficiency; it also changes the direction in which technology develops. When a state perceives the independent consciousness of society as a threat, a technology that could potentially expand social consciousness may instead become an instrument for managing and controlling that very consciousness.

International reports concerning the expansion of digital surveillance in Iran move this concern beyond the theoretical level. In 2025, the United Nations Independent International Fact-Finding Mission on the Islamic Republic of Iran reported an expansion of forms of surveillance and control, particularly against women, following the “Woman, Life, Freedom” movement. The use of cameras, communications data, and digital tools to enforce compulsory hijab illustrates how technology can enter directly into the relationship between the state and citizens’ everyday lives.

The issue, however, extends far beyond compulsory hijab. Once an infrastructure consisting of smart cameras, license-plate recognition, identity databases, telecommunications information, and algorithmic analysis has been established, there is no guarantee that its use will remain limited to a single purpose. The same infrastructure used today to identify a particular social violation can technically be used to identify a protester, labor activist, student, journalist, or political opponent.

From Traditional Repression to the Possibility of Algorithmic Repression

Political repression is not a new phenomenon in the history of the Islamic Republic. What AI can change is its scale, speed, and degree of automation. Traditional security institutions required large numbers of personnel to follow individuals, examine communications, analyze documents, and identify networks of political opponents. Artificial intelligence makes parts of this process automatable. Millions of images can be examined far more rapidly; relationships among individuals can be extracted from network data; enormous volumes of text can be categorized and analyzed; and fragmented datasets can be linked together to construct a more comprehensive picture of the activities of an individual or group.

The central issue, therefore, is not the existence of one particular camera or algorithm, but the possibility of creating a broad architecture of data-driven surveillance in which citizens’ identity, visual, location, communications, and behavioral information are interconnected.

This transformation can be described as a transition from exclusively human repression toward the possibility of algorithmic repression. This does not mean that machines independently make political decisions. Rather, political power combines its historical capacity to control society with the processing power of machines. In this context, AI is not the cause of authoritarianism; it is a multiplier of authoritarian power. A government that was already inclined to monitor its citizens before artificial intelligence can now conduct such surveillance faster, more cheaply, and on a vastly larger scale.

For this reason, the ethical problem of artificial intelligence cannot be resolved merely through technical regulations or recommendations for “responsible use.” As long as the structure of power itself is not accountable, the fundamental question remains: Who watches the watchers?

The Militarization of Artificial Intelligence and the Question of Social Priorities

Another dimension of AI development in Iran concerns its relationship with the security and military establishment. Artificial intelligence has become part of modern warfare throughout the world, with growing applications in drones, image analysis, target recognition, electronic warfare, command systems, cybersecurity, and battlefield intelligence processing. Given Iran’s long-term investment in missile, drone, and cyber capabilities, it is not separate from this global trend.

The issue, however, is not simply that a government employs military technology; virtually every technologically capable state does so. The more fundamental question concerns social priorities and the allocation of resources. In a country facing extensive problems in education, healthcare, water management, energy, the environment, and public infrastructure, society has the right to know how available capital, skilled labor, and computational capacity are distributed—and how much is devoted to public welfare versus security and military institutions.

Here, a distinction must be made between the security of society and the security of the structure of power. The security of the people can mean access to clean water, food, housing, hospitals, education, a livable environment, economic security, freedom from violence, and protection from war. The security of a political system, however, may be interpreted as preserving its monopoly on power, controlling protests, and preventing opponents from organizing.

When these two concepts are merged under the single heading of “national security,” resources allocated in the name of protecting the country may in practice be used to protect the structure of power from society. The debate over military and security applications of artificial intelligence in Iran is therefore ultimately a debate about political accountability.

Two Historical Directions for Artificial Intelligence in Iran

None of this should lead to the conclusion that artificial intelligence is inherently an instrument of repression. The same technology that increases the capacity for surveillance can also expand society’s capacity to acquire knowledge, organize itself, and participate consciously in social affairs. A student in a disadvantaged region can gain access through an AI-based educational system to resources that were previously unavailable. A physician can use advanced diagnostic tools. A farmer can reduce water consumption through data analysis. A researcher can review thousands of academic papers more rapidly. A journalist can analyze large collections of public documents. Citizens can examine budgets, the performance of public institutions, and economic data using tools that were once available only to large organizations.

The fundamental contradiction, therefore, is not between society and technology, but over the ownership, control, and social direction of technology.

Iran faces two different paths. In the first, AI serves the concentration of power: the state possesses increasing amounts of information about the population, surveillance systems become more precise, security institutions acquire greater processing capacity, and citizens increasingly become observable, measurable, and analyzable entities while the structure of power itself remains opaque.

In the second path, the same technology can serve the expansion of public knowledge, universal access to medicine and education, rational management of resources, transparency of public institutions, and greater citizen participation. No technical property inherent in an algorithm guarantees which path will be chosen. Social relations and the structure of power determine the answer.

From this point, we arrive at one of the fundamental principles of a New Social Order in the Age of Consciousness. In an authoritarian structure, the relationship between citizens and power is inverted: the people are transparent to power, while power remains opaque to the people. The government can collect data about citizens, examine their communications, and monitor their behavior, while governmental decisions, contracts, expenditures, the operations of security institutions, and the actual processes of power remain hidden from society. If artificial intelligence develops without transforming this relationship, the imbalance can become dramatically deeper.

A social order based on consciousness must reverse this relationship:

Privacy for citizens; transparency for power.

Personal information, private life, communications, and citizens’ data must be protected. At the same time, public budgets, government contracts, the performance of officials, the allocation of resources, and the decisions of institutions exercising power in society’s name should be transparent to society to the greatest extent possible.

In such a model, artificial intelligence, rather than becoming a tool through which power knows the people more thoroughly, can become a tool through which the people know power more thoroughly and hold it accountable.

Conclusion: The Central Question Is Not the Intelligence of the Machine, but the Consciousness and Power of Society

An examination of the development of artificial intelligence in Iran shows that the country’s situation cannot be explained through the simple binary of “advanced or backward.” Iran possesses significant human capital and a substantial research history, and its society has demonstrated a capacity to absorb and use new technologies. At the same time, weaknesses in computational infrastructure, limited access to capital, sanctions, the migration of specialists, internet restrictions, and, above all, a centralized and security-oriented political structure have prevented this capacity from being freely and broadly mobilized for social development.

From this perspective, the central contradiction of artificial intelligence in Iran is not between “Iran and the West,” or even between “advanced technology and backward technology.” It is between the scientific capacity of society and a concentrated structure of power that seeks to bring technology under its control.

For this reason, the future of artificial intelligence in Iran cannot be separated from the country’s political and social future. Genuine change cannot be achieved merely by replacing one manager, one institution, or even one government with another set of managers if the fundamental relationship among citizens, data, and power remains unchanged.

A future society must decide questions concerning data ownership, the right to privacy, algorithmic transparency, limits on the state’s use of facial recognition, university independence, internet freedom, and public oversight of security technologies. A social revolution in the age of artificial intelligence, if it is to mean something more than the replacement of those who hold power, must therefore place the democratization of knowledge, data, and technology at its center.

Ultimately, a prison can become intelligent. A camera can become intelligent. A missile can become intelligent. A censorship apparatus can operate with the most advanced algorithms available. None of these, by themselves, constitutes social progress.

The true measure of progress in the age of artificial intelligence is not how intelligent the machine has become, but how much freer the human being has become, how much more conscious society has become, and how much more accountable power has become to the people.

The fundamental question for Iran’s future is therefore this: Will artificial intelligence become a technology through which power knows, monitors, and controls the people, or will society be able to transform it into an instrument for expanding knowledge, freedom, participation, and the conscious control of power by the people?

---

Selected References

Blake Atwood, “Artificial Intelligence in Iran: National Narratives and Material Realities,” Iranian Studies, Cambridge University Press, 2025. A study of official narratives of technological independence, AI infrastructure, data centers, the National Information Network, and the material realities of AI development in Iran.

Freedom House, Freedom on the Net 2025: Iran. An examination of internet restrictions, censorship, digital surveillance, and the structure of information control in Iran.

Freedom House, An Uncertain Future for the Global Internet, 2025. An examination of the relationship among artificial intelligence, digital sovereignty, national internet systems, and the expanding surveillance capacity of authoritarian governments.

United Nations Independent International Fact-Finding Mission on the Islamic Republic of Iran, 2025. Reporting on repression and the expansion of surveillance of citizens, particularly women, following the “Woman, Life, Freedom” movement.

Oxford Insights, Government AI Readiness Index. Comparative data on governments’ capacities to develop and deploy artificial intelligence.

This is the complete English version, following the Persian article section by section rather than summarizing it. You can paste it into the new WordPress page now.