Condor’s Ghost – How a Cold War Assassination Program Was Rebranded

On 16 September 1976, a cable (what the Americans call their diplomatic telegrams) went out from Henry Kissinger’s office to U.S. ambassadors in Latin America. It contained a direct instruction: ‘Take no further action’. The message referred to a diplomatic warning the State Department had prepared, one that could have signalled U.S. opposition to the international assassination campaign known as Operation Condor.

That warning was never delivered.

Five days later, a car bomb exploded in Washington, D.C., killing Chilean exile Orlando Letelier and his American colleague Ronni Moffitt. The declassified cables don’t just suggest negligence, they document a specific, high-level decision to stand down just before the killers struck in the heart of the U.S. capital.

The Official Story Has Holes

Most textbooks and government statements put the birth of Operation Condor on 28 November 1975. On that day, senior intelligence officials from five South American countries (Chile, Argentina, Uruguay, Paraguay, and Bolivia) signed a formal agreement in Santiago. The idea, they claimed, was simple… build a regional alliance to combat terrorism and subversion. A kind of Interpol for the Cold War.

That date is still cited in legal rulings and official histories as the beginning of Condor. But it’s not true.

A CIA intelligence report dated 23 June 1976 quietly confirms that the Santiago meeting was not the beginning of anything. According to the report, the core members of the Condor network had already met in ‘early 1974’, in Buenos Aires, to ‘prepare coordinated actions against subversive targets’. That’s almost two years before the so-called founding ceremony.

And they weren’t just planning. They were killing.

In September 1974, Chilean General Carlos Prats and his wife Sofía Cuthbert were murdered in Buenos Aires. A car bomb tore through their vehicle, killing them instantly. Prats had been a loyalist to Chile’s ousted president Salvador Allende and a known opponent of the Pinochet regime. The operation was carried out by Michael Townley, a U.S. citizen working for DINA (Chile’s secret police), with support from Argentine collaborators. It was a textbook Condor mission, cross-border, targeted, and designed to eliminate political dissidents abroad.

This operation took place over a year before Condor was supposedly created.

The official cover story was equally misleading. When confronted by U.S. officials, DINA Director Manuel Contreras insisted Condor’s ‘only purpose is the exchange of intelligence concerning the extremists’. But buried in the declassified documents is a different story.

An FBI cable from September 1976 outlines the existence of a secret ‘Phase III’ of Operation Condor. Under this protocol, special teams could be dispatched to any country in the world to carry out ‘sanctions’, a code word for murder. These weren’t rogue missions. They were part of the structure.

A month earlier, on 3 August 1976, Assistant Secretary of State Harry Shlaudeman sent a memo to Kissinger. It was called ‘The Third World War and South America’. In it, he wrote bluntly that the Southern Cone governments had ‘established Operation Condor to find and kill terrorists… in their own countries and in Europe’.

The distinction matters. If the network was already active, and the U.S. already knew, then the whole idea that Condor’s violence blindsided Washington starts to fall apart.

We are not just talking about an error in paperwork. This is a case where the official founding date conceals a far longer period of coordination, killings, and complicity. The Santiago agreement was not the beginning of Operation Condor. It was just the moment someone decided to give it a name.

Washington’s Two Sets of Books

The official U.S. account of Operation Condor is one of slow realisation. According to that version, it wasn’t until mid-1976 that Washington began to grasp what Condor really was. A joint Latin American security arrangement that had moved from intelligence sharing to overseas assassinations.

But the documents tell a different story. And once you understand what the U.S. already knew, the idea of surprise doesn’t hold up.

Condor’s member states used encrypted teletype machines to send secure messages across borders. The machines were made by a Swiss company called Crypto AG. What the Condor states didn’t know is that Crypto AG had been secretly co-owned by the CIA and the West German intelligence service since the 1950s.

This wasn’t passive ownership. The U.S. intelligence community had deliberately inserted weaknesses into the encryption algorithms used in the Crypto AG devices. That meant that even though Condor operatives believed their communications were secure, the CIA and NSA could intercept and decrypt them in real time. The Condor states called their network ‘Condortel’. The CIA called the backdoor Operation Rubicon.

From the very beginning, U.S. intelligence had the technical capability to monitor Condor’s encrypted messages as they were being sent, including plans, target lists, logistics, the lot. This wasn’t a vague rumour passed up the chain. It was hard data flowing into U.S. hands, directly from the network itself.

That makes the next part harder to explain.

On 10 June 1976, just three months before the Letelier assassination, Henry Kissinger met privately with Admiral César Guzzetti, the foreign minister of Argentina’s new military junta. Guzzetti outlined what he called ‘joint efforts’ with other countries to combat terrorism. Kissinger’s response was not cautious. He said, ‘If there are things that have to be done, you should do them quickly,’ and followed it with, ‘We want you to succeed.’

But while some parts of the U.S. government were offering encouragement, others were raising alarms.

The FBI attaché in Buenos Aires, Robert Scherrer, was documenting Condor clearly and accurately. On 28 September 1976, just days after the Letelier assassination, he sent a cable to Washington describing Condor’s ‘Phase III’ protocol, special teams authorised to carry out assassinations abroad. He specifically flagged the murder in Washington as likely being a Condor operation.

Meanwhile, the CIA was still paying Manuel Contreras, head of Chile’s DINA and one of Condor’s founding architects, as a covert intelligence asset. That relationship continued until 1977, long after the agency knew what Condor was doing and who was directing it.

This wasn’t just inter-agency confusion. It was two tracks running in opposite directions. On one track: U.S. diplomats and FBI officials raising alarms, documenting murders, and preparing formal protests. On the other: U.S. intelligence agencies intercepting encrypted Condor messages, continuing to fund the very people planning assassinations, and providing quiet encouragement at the highest levels.

The question isn’t whether Washington knew. It’s why it kept two sets of books.

The Five Days in September

By August 1976, there was no doubt inside the U.S. State Department about what Operation Condor had become. Multiple intelligence cables, briefings, and internal memos had warned that Condor wasn’t just about sharing information. It was planning assassinations abroad, including in Europe and possibly on U.S. soil.

In response, the State Department did what you would expect from a government that wanted to stop political killings. It drafted a formal diplomatic protest, known as a demarche. The plan was to deliver it to the governments of Chile, Argentina, and Uruguay.

The demarche was carefully worded to express ‘deep concern’ about rumours of assassination plots and to remind the Condor countries that any murder of political exiles on foreign soil would damage their relations with the United States. It was designed to be firm but measured enough to send a warning without provoking a diplomatic crisis.

But when the draft was circulated to U.S. embassies in the region, the reaction was mixed. The ambassadors to Chile and Uruguay pushed back. They warned that delivering the message could strain ties with the military regimes, possibly undermining U.S. influence.

For a few weeks, the demarche hung in the air. It had been approved by senior officials in Washington. The evidence supported it. The threat was real. But nobody had pulled the trigger.

Then came the rescission.

On 16 September 1976, a cable was sent from Kissinger’s travelling party in Lusaka, Zambia. The instruction was simple. ‘The Secretary has instructed that no further action be taken on this matter.’ In other words, the demarche was dead. The warning would not be delivered.

Five days later, on the morning of 21 September 1976, Orlando Letelier was driving to work at the Institute for Policy Studies in Washington, D.C. With him in the car was Ronni Karpen Moffitt, a 25-year-old fundraiser, and her husband, Michael. As they rounded Sheridan Circle, less than two miles from the White House, a bomb planted under the car by DINA agents exploded. The blast killed Letelier and Moffitt. Michael Moffitt survived.

The timeline is stark. A diplomatic intervention designed to prevent exactly this kind of attack was stopped by a direct order from the Secretary of State. Five days later, the attack happened. The question is not whether the U.S. failed to act. The documents show they decided not to.

The Playbook After Condor

Operation Condor officially ended with the fall of the Argentine junta in 1983. The network collapsed along with the dictatorships that ran it. But its core architecture, the model it pioneered, didn’t disappear. It evolved.

Look at the basic structure… The U.S. provides training, technology, and intelligence support to regional security forces. These forces share information across borders and coordinate operations against a designated enemy. The enemy is defined broadly enough to include not just armed groups but political dissidents, journalists, and activists.

In 2000, the U.S. and Colombian governments launched Plan Colombia. Publicly, it was about fighting drug cartels and strengthening democratic institutions. The U.S. provided billions in military hardware, surveillance technology, and training. Colombian forces would use these tools to dismantle narco-trafficking networks.

Seven years later, the Mérida Initiative applied the same model to Mexico. Again, the stated enemy was transnational crime. Again, the solution was U.S.-provided technology and training for local security forces.

The rhetoric had changed. Instead of fighting ‘subversives’, they were fighting ‘narco-terrorists’. Instead of preventing the spread of communism, they were preventing the spread of drug violence. But strip away the language and look at the structure. It’s the same playbook.

The key innovation was updating the definition of the enemy. During Condor, ‘subversive’ was expanded to include nearly anyone who opposed government policy. In the modern era, powerful surveillance tools procured to fight ‘serious crime’ are systematically redirected against journalists, human rights activists, and political opponents.

From Analogue Terror to Digital Spying

The evolution from Operation Condor to modern surveillance programmes isn’t just about new technology. It’s about how the same patterns of abuse persist even as the tools change.

Condor operatives had to physically follow targets, tap individual phone lines, and intercept paper communications. Today’s intelligence services can monitor entire populations through their phones, hack any device remotely, and process millions of communications in real time. But the impunity remains the same.

During Condor, state secrecy and official denial protected operatives from accountability. Governments simply denied the network existed. When evidence emerged, they classified it or destroyed it. Modern surveillance operates in a similar legal vacuum. Governments refuse to confirm or deny their use of tools like Pegasus. The contracts with companies like NSO Group are secret. The legal frameworks authorising their use, if they exist at all, are classified under national security exceptions.

Look at how this played out in Colombia.

Under Plan Colombia, the U.S. helped build sophisticated electronic surveillance systems. The DEA heavily supported the development of Colombia’s ‘Esperanza’ system. The stated purpose was tracking drug shipments and cartel communications.

But in 2009, the ‘Las Chuzadas’ scandal revealed what was really happening. Colombia’s intelligence agency, DAS, had used these U.S.-supplied surveillance systems to illegally wiretap Supreme Court justices, journalists, human rights activists, and political opponents. The technology meant for Pablo Escobar’s successors was turned on anyone who challenged the government.

Mexico followed the same pattern. The Mérida Initiative included funding for advanced surveillance technology. By 2017, forensic analysis confirmed the Mexican government was using Pegasus spyware, purchased from Israel’s NSO Group, to target journalists investigating military abuses and human rights defenders. Once again, tools justified by the war on drugs became weapons against civil society.

This isn’t a bug in the system. It’s a feature. When you hand powerful surveillance tools to security forces with a history of political repression, when you do it with minimal oversight, when the definition of ‘enemy’ is kept deliberately vague, this is what happens. Every time.

Has the School for Dictators Really Closed?

Few institutions embody the Condor legacy like the U.S. Army School of the Americas. For decades, it trained Latin American military officers who later led coups, directed death squads, and operated torture centres. Condor’s key players, from Argentina’s Roberto Viola to Panama’s Manuel Noriega, were SOA graduates.

In 2000, after years of protests, Congress closed the SOA. In 2001, it reopened as the Western Hemisphere Institute for Security Cooperation (WHINSEC) on the same base at Fort Benning, Georgia. Officials insisted this was a genuine reform. The new institution would promote democracy and human rights. Every course would include at least 10 hours of human rights instruction.

Critics saw it differently. They pointed out that WHINSEC opened with many of the same instructors teaching fundamentally similar courses. The core training in military and intelligence techniques continued.

The real test would be transparency. Under the SOA, researchers could track graduates and document their subsequent careers. When SOA alumni showed up leading coups or commanding death squads, the connection was clear.

But since 2004, the Department of Defense has refused to release the names, ranks, and countries of WHINSEC students. Freedom of Information Act requests are denied. Without this basic information, it’s impossible to track whether graduates are later implicated in human rights abuses.

This refusal to provide transparency fuels suspicion. If WHINSEC is truly different from the SOA, if its human rights training is effective, why hide the names of graduates? The addition of human rights courses means little if the institution still teaches the operational skills of repression without fundamentally altering its role in U.S.-Latin American security relationships.

The Files That Remain Secret

For all the revelations about Operation Condor, some of the most critical documents are still locked away. Others were destroyed deliberately. This isn’t a historical clean-up. It’s an ongoing problem.

The most critical missing piece is the archive of ‘Condortel’ intercepts. Given the CIA’s secret ownership of Crypto AG, the NSA would have possessed the capability to decrypt every communication sent through Condor’s network. These files would show real-time operational planning, target selection, and coordination between countries. They would reveal exactly what U.S. intelligence knew and when they knew it.

These intercepts remain classified. Their release would be what investigators call a ‘historical game changer’, potentially revealing the full extent of U.S. foreknowledge and the specific decisions made based on that intelligence.

Key CIA files have been destroyed or remain heavily redacted. The operational file on Manuel Contreras, DINA chief and paid CIA asset, was destroyed in 1991. The agency has acknowledged possessing intelligence on Contreras’s role in ordering the Letelier assassination but continues to withhold this material.

Argentina’s own intelligence archives present another black hole. Despite U.S. declassification efforts that have shed light on the dictatorship, SIDE’s internal operational files remain secret. Investigators and courts must rely on foreign intelligence reports to reconstruct events that happened on Argentine soil. The recent order by the Milei administration to declassify files could change this, but the scope and practical outcome remain to be seen.

Modern surveillance programmes operate with similar opacity. Governments routinely refuse to confirm or deny their purchase and use of spyware. Contracts with surveillance companies are secret. Legal protocols governing their use are classified. This creates a black box around digital surveillance that mirrors the secrecy of the Condor era.

The Core Questions

The evidence leaves us with specific, answerable questions that cut to the heart of the Condor legacy:

• What does the evidence reveal about the gap between the U.S. government’s official timeline of discovering Condor’s assassination plots and its actual, real-time intelligence-gathering capabilities? The Crypto AG operation proves the U.S. could read Condor communications from the beginning. This fundamentally challenges any narrative of gradual awareness.

• Why did Kissinger’s office draft, approve, and then explicitly rescind a diplomatic warning just five days before the Letelier-Moffitt assassination? The cables document a specific decision to cancel preventative action. Who made this decision and why?

• Did the same people who designed Condor simply rebrand their approach for the modern era? The similarities between Condor, Plan Colombia, and the Mérida Initiative are hard to ignore. The same structure, same methods, just different enemies on paper.

• To what extent was the official closure of the SOA and its reopening as WHINSEC a cosmetic rebranding versus fundamental institutional reform? The refusal to provide basic transparency about graduates suggests institutional culture remains unchanged despite formal reforms.

• What would the still-classified Condortel intercepts likely reveal about the day-to-day planning of Condor’s operations? These represent the single most significant information gap and could definitively resolve questions about U.S. foreknowledge and complicity.

The Pattern Continues

Operation Condor wasn’t some Cold War oddity that died with the dictatorships. It was a blueprint. The car bombs became spyware, the torture centres became server farms, but the basic recipe stayed the same. To give local forces powerful tools, let them decide who the enemy is, and make sure nobody can prove you knew what would happen next.

The Letelier-Moffitt assassination happened because the U.S. government chose not to prevent it. That choice was documented in cables and memoranda that were never meant to be read by the public. Today, similar choices are being made about surveillance technology that enables political repression across Latin America.

The difference is that today’s choices are happening in real-time, and the digital tools being deployed are more powerful than anything the Condor states could have imagined. The question isn’t whether this technology will be abused. The evidence shows it already is. The question is whether we’ll keep pretending to be surprised when it happens.

The files that remain secret contain the answers we need. Until they’re released, the afterlife of Operation Condor will continue unexamined, and the pattern will persist. Some lessons from the Cold War were never learnt because they were never fully exposed. That’s not an accident. It’s policy.

Sources

Sources include: declassified U.S. government records from the Central Intelligence Agency, Department of State, and Federal Bureau of Investigation, including intelligence memoranda on early Condor planning, State Department cables ordering the rescission of diplomatic warnings (September 1976), and FBI attaché reports from Buenos Aires identifying Condor’s assassination protocols; judicial records and verdicts from the landmark “Condor Trials” in Argentina (2016) and Italy; foundational documents from the Paraguayan “Archives of Terror” discovered in 1992, providing the first internal evidence of the network’s structure; reporting from The Washington Post and German broadcaster ZDF on the CIA’s co-ownership of encryption firm Crypto AG; U.S. legislative records concerning the closure of the School of the Americas (2000) and the establishment of its successor, WHINSEC; official documentation and U.S. congressional reports on Plan Colombia and the Mérida Initiative; forensic analysis from digital rights organisations on the use of Pegasus spyware in Mexico; and extensive archival work and contextual analysis from non-governmental organisations, principally the National Security Archive.