The Network: The CIA in China, and the Night It Died
Does the CIA really operate inside China? The proof isn't speculation — it's in American court documents, a 19-year sentence, and the agency director's own admission. The story of the network, the 2010-2012 catastrophe that destroyed it, and the decade of rebuilding.
By MyAudioBooks.ai ·
Listen free: The Network: The CIA in China, and the Night It Died
The question is the kind that sounds like it belongs to conspiracy forums: does the Central Intelligence Agency actually run operations inside China? Not near China, not against China's networks in America, but inside — agents on the ground, recruited sources in ministries and military commands, the whole human machinery of espionage operating within the borders of the world's most surveilled state. The answer is yes, and the remarkable thing is where the proof lives: not in leaked Chinese accusations or internet speculation, but in American court documents, in the admissions of the agency's own director, and in the official record of the worst American intelligence catastrophe of the twenty-first century — a catastrophe so complete that for two years, between late twenty ten and the end of twenty twelve, the Chinese government hunted down the CIA's entire agent network inside China and destroyed it, source by source, until somewhere between eighteen and thirty people who had been feeding secrets to American intelligence were in prison or dead. The United States does spy inside China. It did then, it does now, and the evidence for both claims is sitting in documents anyone can read — most of them written by the United States government itself. This is the story of the network, the night it died, and the decade of rebuilding that its own director confirmed from a podium.
At My Audio Books dot A I, you can create your own audiobooks from prompts, turn your documents into audio, all with one subscription, and store your items in your own personal library.
Start with the catastrophe, because it is the documented fact around which everything else in this story turns. Beginning in late twenty ten, officers of China's Ministry of State Security — the MSS, the civilian intelligence service that runs Beijing's counterintelligence — began arresting the CIA's sources inside China with a precision that told the agency immediately that this was not bad luck. One by one, across ministries, military offices, and provincial governments, the people who had spent years passing information to their American handlers — the case officers who run sources — were identified, arrested, and in a number of cases executed, some of them reportedly in the courtyards of the buildings where they worked, as a lesson to their colleagues. The reporting that reconstructed the disaster, years later, put the toll at eighteen to twenty sources killed or imprisoned in roughly two years, with later accounts running the number as high as thirty. The CIA's human network inside China — painstakingly built over decades, at the cost of careers and fortunes and lives — was, in the words of the officials who finally spoke about it, crippled: not degraded, not diminished, but functionally annihilated, leaving the United States nearly blind inside its most important adversary at the exact moment China was becoming one.
Why the network died is the second document trail, and it is the part that makes this a story about the CIA's own failures rather than merely China's competence. Two explanations run through the record, and the honest answer is that both are probably true. The first is technical, and it is devastating: the communications system the agency used to talk to its sources in China — a covert digital channel built for short, secure exchanges and originally developed for use in the Middle East — turned out to have a fatal flaw: once you knew it existed, its architecture made the agency's sources traceable, and the Chinese found it. When the same system was penetrated in Iran and its sources there were rolled up, the flaw was exposed as systemic — and the Chinese network, running on the same machinery, was already dying. The second explanation is human, and it has a name and a docket number: Jerry Chun Shing Lee, a former CIA case officer who left the agency in two thousand seven, took classified notebooks with him — handwritten records of the names, locations, and operational details of CIA assets and covert employees, including people in China — and was later found to have been meeting with Chinese intelligence officers who paid him for what he knew. The Justice Department indicted him in twenty eighteen and sentenced him to nineteen years in prison in twenty nineteen, on conspiracy to deliver national defense information to a foreign power. Whether Lee's notebooks were the map that led the MSS to the network, or the communications flaw was the map and Lee merely confirmed it, the court record and the reporting have never fully settled — but the documentary proof of both failure modes exists, in an American courtroom file and in the agency's own postmortem reporting, for anyone who doubts the catastrophe was real.
At My Audio Books dot A I, you can listen to this story and thousands of others that explore the hidden science and mechanics behind the headlines.
The confession from the podium is the third document, and it is the one that answers the question this article opened with in the present tense. In July of twenty twenty-three, the Director of the Central Intelligence Agency, speaking at the Aspen Security Forum, said something no sitting director would have said a decade earlier: that the agency had made progress rebuilding its human intelligence network inside China — confirming, in a single sentence, three facts at once: that the network had been destroyed, that rebuilding it was a priority, and that the work of rebuilding — running agents inside China, today — was underway. Directors do not make such statements carelessly; the disclosure was a deliberate signal, part deterrence theater and part pride, but its content was factual: the CIA operates inside China, and the director of the CIA said so on the record. The fourth document trail runs from the other side of the wall: China's own security services have spent the years since telling on themselves, in their way — the MSS now maintains public channels through which it announces espionage cases, warns citizens about foreign recruitment, and names the CIA as the adversary in its public campaigns, each announcement an inadvertent confirmation that there is something inside China for them to hunt. And the fifth trail is the oldest: the agency's own declassified history documents decades of operations inside China going back to the nineteen fifties — the teams, the drops, the lost flights — establishing that what was destroyed in twenty twelve and rebuilt after was not an aberration but the continuation of the CIA's oldest and hardest mission.
Now the harder question, the one the documents point at but cannot fully answer: what does running agents inside China actually mean in twenty twenty-six, and what is it for? The mission has changed shape since the network died. The classic model — the recruited official passing documents to a case officer — is nearly suicidal in the most surveilled society on Earth, where ubiquitous cameras, mandatory apps, and a counterintelligence service with unlimited domestic reach make the face-to-face tradecraft of the Cold War close to impossible. The rebuilt network, by every indication in the public record, leans on the twenty-first century's answers: longer-range recruitment of Chinese nationals while they are abroad, virtual handling through the same kinds of secure digital channels whose earlier failure cost the network its life — rebuilt, one hopes, better — and technical collection woven so deeply into the human mission that the line between a spy and a sensor blurs. And the purpose has sharpened: the documents the network exists to steal are no longer general political intelligence but the specific secrets of a peer military — weapons programs, command intentions, and above all the plans for Taiwan, the contingency this channel covered in the summit story last week, for which American intelligence needs to know not what Beijing says but what Beijing intends.
The strongest case that the catastrophe was a one-time disaster, now repaired — stated at full strength, because the agency's own confidence is built on it — begins with the observation that the failure's causes were identified and, in principle, fixed: the flawed communications system was retired and replaced, the tradecraft for the most dangerous environment on Earth was rebuilt from first principles, and the mole question — whether a traitor inside the service had given the network away — was answered in a federal courtroom with a nineteen-year sentence that closed the book on Lee. The rebuilding has had a decade to mature under directors who made China the agency's top priority, with resources to match, and the director's own public statement — rebuilding, progressing — is the institutional assessment that the network is back. And the counterintelligence asymmetry cuts both ways: the same era saw Chinese intelligence operations inside the United States exposed at a remarkable rate — the indictments of MSS-linked officers, hackers, and recruiters in American courts read as a mirror of the Lee case — suggesting that both services penetrate each other, and that the American service is hardly the only one taking losses in the war between the secret worlds.
And the strongest case against the comforting version — stated with the precision the stakes demand — is that the conditions that destroyed the network have gotten worse, not better, and the rebuilt one operates against a harder wall than the old one ever faced. The surveillance state the MSS commands today is beyond anything the twenty-ten network had to evade: a domestic apparatus of cameras, facial recognition, mandatory digital identity, and data-fusion that makes the physical acts of espionage — the meetings, the drops, the brush passes, the fleeting in-person exchanges of the trade — visible by default rather than by investigation. The rebuilding is also unverifiable from outside: the director's claim of progress is self-reported, the agency has institutional reasons to project recovery, and the last time the agency believed its Chinese network was healthy, it was already dying without knowing it — which is the deepest lesson of the catastrophe: you do not know your network is compromised until the arrests begin. And the asymmetry of consequences remains brutal: an American intelligence failure in China costs the sources their lives and the country its eyes, while a Chinese failure in America costs Beijing a few indicted officers and a diplomatic note — an imbalance that makes every source recruited in China a person betting their life on a system that has already failed once in living memory.
Three developments would disprove or confirm the state of the rebuilt network in the years ahead, and each is visible in public documents. First, the MSS's own announcements: China's security service publicizes its counterintelligence victories, and any future wave of espionage-case announcements naming American networks will tell us — as the arrests told the agency in twenty ten — whether the rebuilt network has been found; silence in that channel is the closest thing to evidence of health we will ever get. Second, the American court record: espionage cases move both ways, and future indictments — of Americans spying for China, or of Chinese officers running operations against the rebuilt networks — will keep mapping the war between the services in the only fully public ledger that exists. Third, the Taiwan file: the quality of American foresight in any future cross-strait crisis — whether Washington reads Beijing's intentions early or is surprised — will be the rebuilt network's ultimate report card, graded in public, in real time.
It is worth saying what this article has not claimed. It has not claimed any specific current operation; the rebuilt network's details are secret, and the article has relied only on public documents, named as such. It has not claimed the Lee case caused the catastrophe alone; the communications-flaw explanation is given at equal weight, as the record requires. It has not claimed the rebuilding has failed or succeeded; the only honest verdict is that it is unverifiable, and the article says so. And it has not claimed espionage is unique to the United States; the mirror cases of Chinese operations in America are documented here at length. The claim here is narrower and fully proven by the record cited: the CIA has run, and runs, operations inside China; the proof is in American court documents and the agency's own public statements; and the night the old network died remains the worst American intelligence catastrophe of the century, documented by its own side.
Which returns to the question that sounds like a forum post and turns out to be a court record. Does the CIA operate inside China? The Justice Department has a file that says it did. The agency's director has a podium transcript that says it does. And somewhere in Beijing today, the Ministry of State Security is running public campaigns warning its citizens about the foreign intelligence service that recruits them — which is the third government, in its way, confirming the same answer. The conspiracy theory is wrong about the details, as conspiracy theories always are. About the fact, it is simply, documentedly, correct.
At My Audio Books dot A I, you can create fiction, non-fiction, and turn your documents into audio, all stored in one place with a single subscription — plus get instant access to thousands of audiobooks and deep-dive investigations. Learn more today at My Audio Books dot A I.