A recent investigative report by Associated Press makes a provocative claim: that U.S. technology companies helped enable China’s modern surveillance system, laying the groundwork for mass monitoring, detention and social control.
On the surface, this appears to be a rare moment of Western self-criticism. American firms sold the tools. The United States failed to restrain them. China used the technology in ways that raised serious human-rights concerns.
But read carefully and a different pattern emerges.
Despite pointing the finger at Silicon Valley, the article ultimately frames China as the moral endpoint of the problem. The U.S. is portrayed as careless, commercially driven or naïve. China is portrayed as uniquely abusive, uniquely dangerous and uniquely illegitimate.
That framing deserves scrutiny, not because surveillance cannot be misused in China, but because the article collapses context, exaggerates capability and avoids comparison, turning a global structural issue into a China-specific morality tale.
China’s national public-security digitisation effort, often referred to as the “Golden Shield,” began in the late 1990s. This is not disputed. What is often omitted is why.
The late 1990s and early 2000s were a period of extraordinary disruption inside China:
• Massive rural-to-urban migration
• Weak local governance capacity
• Rising crime
• The rapid arrival of the internet and mobile communications
At the same time, the global security environment was shifting. After 9/11, Western governments dramatically expanded surveillance, intelligence fusion and pre-emptive security doctrines. The U.S. Patriot Act, NSA bulk collection programs and predictive policing models were not fringe experiments, they became standard tools of governance.
China did not invent surveillance modernity. It adopted it, as nearly every industrialised state did, under conditions of rapid digitisation and heightened security anxiety.
One of the most emotionally charged elements of the AP article is its focus on “predictive policing”, the idea that algorithms can identify risk before crimes occur.
But predictive policing:
• Was pioneered and popularised in the United States
• Has been trialled extensively by U.S. police departments
• Has been criticised by U.S. civil-rights organisations for bias and overreach
• Is used by immigration authorities, border agencies and counter-terror units across the West
The ethical risks of these systems are well documented. They are not Chinese discoveries, they are features of modern data-driven governance.
Framing predictive policing as something that becomes inherently abusive only when deployed by China is intellectually dishonest.
The article repeatedly leans on imagery of omnipresent cameras to suggest a kind of all-seeing state.
This is where narrative overtakes reality.
Yes, China has a large number of cameras, particularly in cities. But no state, including China, has the technical, human or logistical capacity to actively monitor 1.4 billion people.
Cameras are largely passive collection tools, not instruments of constant human observation. Most footage is never viewed. Artificial-intelligence systems operate on event-based triggers, not continuous oversight. Storage, processing limits and staffing constraints alone make the idea of total population monitoring impossible.
This is not unique to China. London is among the most camera-dense cities in the world. U.S. cities combine public surveillance with vast private camera networks. Japan, South Korea and Singapore all rely heavily on camera-based public-safety systems.
In China, cameras are primarily used for:
• Traffic enforcement
• Crime investigation
• Missing-person searches
• Emergency response
• Crowd and transport safety
That does not mean misuse never occurs. It does mean that portraying China as operating a real-time “digital prison” is an exaggeration, not an operational description.
Any serious discussion of Xinjiang that omits the security background is incomplete.
From the 1990s through the mid-2010s, Xinjiang experienced repeated violent incidents, including:
• Mass-casualty knife attacks
• Bombings
• Vehicle attacks
• Assaults on civilians, markets, police stations and transport hubs
Many of these incidents were linked to global extremist networks and authorities documented cases of foreign-fighter travel and return. These events occurred during the same global period in which the United States, Europe and others dramatically expanded surveillance and preventive security measures in response to terrorism.
No state, none, treats sustained terrorism as a routine policing matter.
Whether China’s response in Xinjiang was proportionate, excessive or flawed is a legitimate subject of debate. But portraying the region as a surveillance-driven repression project in a vacuum, stripped of its security history, is misleading.
It transforms a complex counter-terrorism challenge into a moral tale.
Since the late 2010s, large-scale terrorist attacks in Xinjiang have sharply declined and the mass-casualty incidents that once defined the region have not occurred since. Public spaces, transport hubs and markets operate without the sustained violence seen in earlier years. Whether one agrees with every method employed or not, the security outcome itself is not seriously disputed: the campaign succeeded in restoring basic public safety and preventing further large-scale attacks.
The article repeatedly implies that American technology formed the foundation of China’s surveillance apparatus.
This is a partial truth presented as a total one.
American firms sold servers, databases, networking equipment, chips and enterprise software, all general-purpose technologies sold globally to governments and corporations alike. They did not design China’s policing ideology, legal framework or deployment strategy.
Crucially:
• The architecture
• The policy logic
• The scope of deployment
• The decision to scale
All were Chinese state choices, increasingly executed by Chinese companies, especially after 2015–2019 when sanctions accelerated domestic substitution.
Even the AP article acknowledges that U.S. sales slowed while China continued and adapted. That alone undermines the claim that Silicon Valley “built” the system in any meaningful ongoing sense. Infrastructure is not intent.
Perhaps the clearest indicator of narrative bias is what the article does not do.
It does not apply the same scrutiny to:
• U.S. mass surveillance programs
• Western predictive policing failures
• Border surveillance regimes
• Intelligence-policing fusion in democratic states
Instead, China becomes the example to avoid, not surveillance itself.
The implied lesson is not “these systems are dangerous everywhere,” but “this is what happens when China gets them.” That distinction matters.
The uncomfortable truth is this:
Modern surveillance systems emerge where state power, commercial technology and security fear intersect.
That intersection exists in China, the United States and Europe and in any society where states govern large, complex populations under conditions of risk.
China is not an outlier in direction. It is an outlier only in political structure and that difference is used to moralise outcomes rather than analyse systems.
Blaming China for becoming a “surveillance superpower” avoids the harder question: Why did every technologically advanced state move in the same direction?
The AP investigation raises legitimate questions about corporate ethics, export controls and the global spread of surveillance technology. Those questions deserve serious debate.
What it does not justify is turning China into the villain of a story that is fundamentally about modern governance in the digital age.
China did not invent surveillance modernity, predictive policing, or the data-fusion and algorithmic risk-scoring systems that now characterise modern governance. It adopted tools already normalised elsewhere and applied them within its own political system.
If surveillance is the problem, China is not the exception.