The question of whether Iran and China are jointly conducting spying operations has gained a new technological dimension. Recent findings from Anthropic indicate that state-linked actors in both countries have used artificial intelligence to support surveillance activities, including efforts involving dissidents, activists and other targets. But the evidence currently available does not establish that China and Iran are operating a single, coordinated espionage network. What it does reveal is something potentially more consequential: authoritarian states with different geopolitical interests are increasingly discovering that AI can dramatically lower the cost of surveillance.
This distinction matters. It would be premature to describe every parallel Chinese and Iranian operation as a joint intelligence campaign. Yet the growing use of the same advanced technologies for monitoring people raises a broader question: Are China and Iran moving toward a shared model of AI-enabled surveillance, even without formally sharing intelligence operations?
AI Is Becoming a New Tool of State Surveillance
Anthropic’s September 2026 threat report provides an important warning about how quickly artificial intelligence is moving from experimental technology into the machinery of state surveillance.
According to reporting on the findings, government-linked actors in China, Iran and countries in West Africa used Anthropic’s Claude models in surveillance-related activities. The reported operations included collecting and processing information, identifying targets and monitoring people such as dissidents, journalists, activists and political figures. Anthropic says it disrupted the operations and banned accounts connected to the misuse.
The significance goes beyond the individual cases.
Traditional surveillance requires large teams of analysts to collect information, translate documents, examine social-media activity, identify relationships and produce intelligence assessments. AI can automate substantial portions of that process.
That means governments do not necessarily need dramatically larger intelligence agencies to monitor more information. They can make existing surveillance systems faster and cheaper.
This is where the China-Iran question becomes strategically important.
China and Iran: Cooperation or Parallel Surveillance?
There is currently a major difference between using similar technology and jointly conducting espionage operations.
The available evidence supports the first conclusion much more strongly than the second.
China and Iran already maintain a strategic relationship covering energy, trade, technology and diplomatic cooperation. Their governments also share opposition to several aspects of US-led international pressure. Beijing has become an important economic partner for Tehran, while both countries have sought to reduce their vulnerability to Western sanctions and technological restrictions.
But strategic alignment does not automatically mean that Chinese and Iranian intelligence agencies are running common surveillance missions.
There is no publicly established evidence in the Anthropic findings that Beijing and Tehran jointly planned the reported AI surveillance operations as one coordinated campaign.
The more defensible conclusion is therefore more subtle: China and Iran appear to be part of a wider global trend in which governments are exploiting AI to expand surveillance capabilities.
That distinction should not weaken the warning. It makes the development more significant because the technology can spread without requiring a formal intelligence alliance.
China’s Surveillance Model Has Already Gone Global
China’s experience with digital surveillance gives it an important advantage in the emerging AI-security environment.
For years, Chinese authorities have invested heavily in facial recognition, data analysis, internet monitoring and automated security systems. AI can now connect many of these capabilities into a much more powerful analytical system.
The danger is not simply the camera or the database. It is the ability to connect information.
A government can potentially combine social-media activity, public records, travel information, communications metadata, photographs and online behavior and then use AI to identify patterns that would be difficult for human investigators to detect manually.
That creates a fundamentally different surveillance environment.
China’s technological development has therefore become relevant beyond China’s borders. Chinese surveillance technologies and security systems have been exported or marketed internationally, while Beijing’s broader digital-governance model has attracted attention from governments seeking greater control over information and political activity.
This is why the Iran-China relationship deserves scrutiny even when direct operational cooperation cannot be proven.
Iran Has Its Own Reasons to Embrace AI Surveillance
Iran also has strong incentives to use artificial intelligence for domestic intelligence.
Iranian authorities have long faced political opposition, demonstrations, dissident networks and pressure from overseas opposition groups. Digital communications have become increasingly important to political organizing, particularly when conventional media and public political activity face restrictions.
AI potentially gives Iranian security agencies a new way to process enormous quantities of digital information.
Instead of manually examining thousands of accounts, documents or communications, AI systems can help classify information, identify relationships and prioritize individuals for further investigation.
The danger is particularly acute for journalists, activists and dissidents.
An AI system does not necessarily need to “understand” politics in the human sense. It only needs to identify patterns associated with people whom authorities consider politically significant.
That can transform surveillance from a labor-intensive activity into a scalable technological process.
The Most Important Development May Be the Cost of Spying
Perhaps the most important lesson from the Anthropic findings is economic.
AI can make surveillance cheaper.
This could fundamentally change the balance between governments and civil society.
Historically, extensive surveillance required substantial personnel, technical infrastructure and intelligence resources. Advanced AI can reduce some of those requirements.
A relatively small team could potentially process enormous quantities of publicly available information.
That creates a troubling possibility: governments do not have to choose between monitoring a small number of high-priority targets and conducting broad information collection. AI makes it easier to do both.
Anthropic’s threat-intelligence chief has emphasized that AI can make state-sponsored surveillance more cost-effective and efficient, even when it does not necessarily expand the number of targets.
The distinction is important because efficiency itself can become a form of power.
Could China and Iran Share Surveillance Technology?
This is where the geopolitical question becomes more complicated.
Even if there is no evidence of a formal China-Iran spying operation, cooperation could occur through technology, expertise, commercial relationships or indirect knowledge transfer.
China has a large domestic technology industry and significant experience in AI-powered security systems. Iran, meanwhile, has a strong incentive to obtain technologies that can help it overcome sanctions, censorship challenges and information restrictions.
A future relationship could therefore involve more than traditional military cooperation.
It could include:
AI surveillance tools, cyber capabilities, data-analysis systems, facial recognition, automated monitoring and information-management technologies.
Such cooperation would not necessarily require Chinese intelligence officials and Iranian security officials to sit together and design a joint operation.
Technology itself can create interoperability.
The Bigger Threat Is the AI Surveillance Ecosystem
The emerging threat should therefore not be reduced to the question of whether China and Iran are spying together.
The bigger question is whether an international AI surveillance ecosystem is developing in which governments learn from each other’s methods and acquire increasingly sophisticated tools.
The Anthropic report is particularly significant because the cases it describes involve more than one region and more than one type of state actor. The same underlying technology can potentially be adapted to different political environments.
That makes AI different from many traditional surveillance technologies.
A sophisticated surveillance platform once required specialized hardware and highly trained personnel. Modern AI can increasingly be accessed through software, cloud infrastructure and commercially available models.
The technology can therefore spread much faster.
China’s AI Ambitions Make the Issue Even More Sensitive
The surveillance issue also comes at a moment when the United States and China are engaged in an increasingly intense technological competition.
Washington has recently accused Chinese AI companies of conducting large-scale model distillation to reproduce the capabilities of advanced American AI systems. China has rejected the accusations and argued that distillation is a legitimate technical method.
At the same time, the United States and China are preparing discussions on AI safety, including concerns over AI-enabled cyberattacks.
This creates an extraordinary contradiction.
The world’s major powers are simultaneously competing over AI capabilities and trying to establish rules to prevent those same capabilities from becoming dangerous.
Meanwhile, governments are already experimenting with AI for surveillance, cyber operations and influence activities.
Is This a New Form of Digital Authoritarianism?
The development could eventually produce something more powerful than conventional surveillance.
Imagine a system capable of continuously monitoring online conversations, identifying politically sensitive language, mapping relationships between individuals, translating foreign-language material, detecting emerging protest networks and generating intelligence reports.
That would not simply be surveillance.
It would be predictive political monitoring.
The system could potentially identify a person as a future political threat before that person had committed any conventional offense.
This is where civil-liberties concerns become particularly serious.
AI systems can also make mistakes. Algorithms can misinterpret sarcasm, political criticism, cultural references or innocent associations. If governments treat algorithmic assessments as intelligence rather than probabilistic analysis, innocent people could become targets.
The problem becomes even more dangerous when there is no independent judicial oversight or transparent mechanism for challenging an AI-generated classification.
China and Iran May Not Need a Formal Spy Alliance
The strongest conclusion from the current evidence is therefore not that China and Iran have created a joint espionage organization.
There is insufficient public evidence to make that claim.
The more important possibility is that China and Iran, along with other governments, are entering the same technological era in which artificial intelligence makes surveillance faster, cheaper and more scalable.
Their political interests may overlap.
Their security institutions may learn from similar technologies.
Their technological cooperation could deepen.
And their governments may increasingly view AI as a strategic instrument for controlling information and identifying perceived threats.
That could produce many of the practical consequences of cooperation without requiring a formal intelligence alliance.
The West Faces a New Surveillance Challenge
Western governments should therefore be careful about treating AI surveillance simply as another cybersecurity problem.
The challenge is broader.
It involves human rights, privacy, technology exports, artificial intelligence regulation, cybersecurity and national security simultaneously.
Europe in particular faces a difficult balancing act. It wants to remain competitive in artificial intelligence while also establishing strong protections for privacy and fundamental rights.
The spread of AI surveillance abroad could eventually create another problem: technologies developed for legitimate commercial purposes could be repurposed by governments with very different standards of political accountability.
That means AI companies increasingly have to consider not only what their technology can do, but who can use it and for what purpose.
The Real Question Is Who Controls AI
The debate over China and Iran’s alleged surveillance cooperation ultimately points toward a much bigger geopolitical struggle.
The future contest may not simply be between the United States, China, Iran, Russia or Europe.
It may be between different models of who controls artificial intelligence.
If AI becomes primarily a tool of state surveillance, governments could acquire unprecedented capabilities to monitor citizens, political opponents and diaspora communities.
If AI remains subject to strong safeguards, independent oversight and enforceable human-rights standards, the technology could instead become a powerful tool for economic development and public services without becoming an instrument of political repression.
The Anthropic cases provide an early warning that this battle has already begun.
Joint Spying Is Not Yet Proven, But the Risk Is Real
So, are Iran and China jointly conducting spying operations?
Based on the publicly available evidence, the answer is not proven.
What is increasingly clear, however, is that both countries are operating within a rapidly expanding world of AI-enabled surveillance. Anthropic’s findings demonstrate that governments can use advanced AI to automate parts of intelligence and monitoring operations, including targeting dissidents and political actors.
The greatest danger may therefore not be a secret China-Iran espionage alliance.
It may be the emergence of a world where governments no longer need enormous intelligence organizations to conduct sophisticated surveillance.
AI could make spying cheaper, faster and more scalable—and that could fundamentally change the global balance between state power and individual freedom.
