Surveillance State
A surveillance state is a government that uses mass data collection, AI monitoring, and population tracking to achieve total visibility over its citizens. Professor Jiang argues that this is the true end-goal of AGI: not a robot apocalypse but the creation of an infrastructure that monitors every person in the world in real time — replacing consent-based governance with a system where non-compliance is impossible because it is always detected.
The concept of the surveillance state is older than AI. Jeremy Bentham's Panopticon (1791) described a prison designed so that inmates could always be observed but never know when they were being watched — producing self-policing behavior. Michel Foucault extended this to society as a whole: modernity creates disciplined subjects by making everyone feel permanently visible to authority. Surveillance does not require constant watching; it requires the belief that one might always be watched.
Modern surveillance states combine this psychological dynamic with actual technical capability. China's Social Credit System — drawing on facial recognition cameras, mobile payment data, travel records, and social media activity — is the most fully developed example. Citizens' scores rise and fall based on monitored behavior, restricting access to transport, education, and employment for low scorers. The system is not primarily punitive; it is self-disciplining. People change behavior when they know the data is being collected.
Professor Jiang's analysis of Operation Stargate connects US AI investment directly to surveillance ambitions. The stated justification is national security and economic competitiveness. The actual architecture — massive data centers processing population-scale information — is dual-use infrastructure that can serve corporate advertising or state surveillance interchangeably. The same facial recognition trained on Chinese school children and the same LLMs trained on internet text are the tools of both commercial AI and state monitoring.
The deeper argument is that AGI is not possible without a surveillance state. Supervised machine learning requires clean, labeled data at massive scale. The only entities capable of generating this data are governments that can compel data collection. American AI companies work with Chinese data collection operations because American privacy law prevents the same surveillance at home. The surveillance state is not the dystopian consequence of AI; it is the precondition for it.
Related articles
Related terms
Frequently asked questions
What is a surveillance state? ＋
A surveillance state is a government that uses mass data collection — cameras, AI, financial records, communications monitoring — to track its entire population continuously. The goal is not to catch criminals after the fact but to make non-compliance impossible by making everything visible. China's Social Credit System is the most developed current example: behavior is scored in real time, and scores determine access to transport, education, and employment. Professor Jiang argues that the true purpose of AGI is to build a global version of this system.
How does AI enable the surveillance state? ＋
AI dramatically lowers the cost of mass surveillance. Facial recognition identifies individuals from camera feeds automatically. LLMs analyze communications at scale. Pattern recognition flags anomalous behavior without human review. Without AI, full-population surveillance requires enormous numbers of human analysts. With AI, a small team can monitor a city. Operation Stargate — the $500 billion US government AI investment — builds the data center infrastructure that makes this technically feasible, regardless of whether that is its stated purpose.