Helms, Edwards and the September 18 Plan for Chile

By Desmond Okafor · September 24, 2026

A CIA source's identity was botched by censors in a memo concerning the Chilean political crisis, kept secret until it reached the State Department's public files.

Chileans in 1970 faced the sudden possibility of a Marxist government that would dismantle their economy and civil liberties. For the men in Washington, the stakes were the loss of a strategic Southern Cone stronghold to Soviet influence. The record of how the United States managed this crisis is a study in the distance between the ground truth and the official transcript. Our window into this meeting is narrow, as what survives is a catalogue description rather than the record's own pages.

On September 18, 1970, a meeting took place that was recorded as a "Discussion of the Chilean Political Situation" (CIA Memorandum of Conversation). The participants were Richard Helms, a key participant in the documentation produced and censored by the CIA, and a man named Edwards, an intelligence source for the CIA whose contributions focused on the Chilean political situation. This was not a casual exchange. It happened in the wake of the September 4 election, where Salvador Allende had secured a plurality of the vote, threatening to pull Chile into the orbit of Moscow.

"Discussion of the Chilean Political Situation"

In the world of tradecraft, the title of a document is often the only thing that survives the shredder. The record identifies this session as a conversation between the agency's top officer and a source. According to the public record, Chile's government was a representative democratic republic, but that stability was precisely what the U.S. feared was evaporating. Helms was not there to gather raw data; he was there to synthesize it into a narrative that could be sold upward.

Edwards provided the conduit. He was the eyes and ears on the ground, the man whose access allowed the agency to peer into the machinery of the Chilean state. The record shows that Edwards provided intelligence contributions specifically related to the political situation, acting as the bridge between the street-level anxiety in Santiago and the policy desks in Langley. When a source like Edwards meets with the head of the agency, the objective is rarely discovery. It is confirmation.

Document imagery from nsarchive.gwu.edu From the files: nsarchive.gwu.edu

Edwards and the Black Pen

There is a telling failure in how this document was handled by the agency's censors. The record shows that "CIA censors redacted Edwards' name in parts of the document to protect his intelligence contribution, but left it uncensored in places, including the title."

This is a classic failure of internal discipline. In the haste to move the record or the sloppiness of a clerk, the very identity the agency sought to shield was left staring the reader in the face. If the identity of the source is inconsistent across a single piece of paper, it suggests a breakdown in the chain of custody. The pattern suggests that by the time this memorandum was being processed for long-term storage, the urgency of protecting Edwards had been superseded by the bureaucracy of archiving.

There is a sharper irony here. The agency's attempt to shield Edwards was not a gesture of kindness, but a requirement of the trade. A burned source is a useless source, and a burned source in a place like Chile is a liability. The fact that the name remained in the title while being blacked out in the body creates a map of the agency's priorities: the administrative label mattered more than the man's safety. The desk's reading is that the inconsistent redaction of 'Edwards' suggests the source's identity was either already compromised or had become less sensitive than the specific nature of his access by the time the censors got to the page.

Foggy Bottom's Filter

The journey of this document does not end at Langley. The record shows that this "transcribed version" eventually appeared in the U.S. State Department's Foreign Relations of the United States series. This transition—from a CIA internal memorandum to a State Department publication—is where the real editing happens.

The U.S. State Department archived and published the history of political discussions concerning Chile, but the version they published was a transcription of a CIA original. When a document moves from the clandestine service to the diplomatic service, it is scrubbed of its teeth. The public record tells us the State Department is responsible for the nation's foreign policy and relations, but in this case, it acted as the final filter for the agency's history.

If the shape of this file is what it appears to be, the "transcribed version" is not a mirror of the original conversation; it is a curated echo. The desk's reading is that the transition from an internal CIA memorandum to a State Department publication was used to scrub the record of transactional costs. Intelligence sources like Edwards do not work for patriotism alone. There are payments, promises, and favors. None of those operational costs appear in a transcribed version destined for the archives. The money is always the first thing to be redacted.

September 18, 1970

The date is the most critical piece of evidence. September 18 comes exactly two weeks after the election and just before the Chilean Congress was set to vote on the presidency. This was the window of maximum volatility.

While the document is titled a "Discussion," the pattern suggests that this was not a genuine intelligence gathering session. The hierarchy is too steep. Richard Helms was the Director of Central Intelligence; Edwards was a source. A Director does not "discuss" with an asset; he conducts a read-back. The desk's reading is that this conversation was a confirmation of existing policy—a way for Helms to ensure that the intelligence being fed to the White House aligned with the desired outcome.

Furthermore, the record is conspicuously silent on the presence of any White House principals. In a high-priority national security crisis, the absence of the executive branch in a CIA-authored memorandum is a loud silence. The pattern suggests that the executive branch's direct 'Track II' orders—the covert efforts to provoke a military coup—were intentionally omitted from this specific channel to preserve plausible deniability. The memorandum records the intellectual justification for intervention, providing the 'why' while carefully erasing the 'how.'

The desk's reading is that this document serves as the polished façade of a much uglier operation. The gaps in the record—the missing payments to Edwards, the inconsistent redactions, and the absence of White House directives—are not accidents. They are the boundaries of what the U.S. government was willing to admit it had done. A full release of the original, un-transcribed memorandum would likely show a conversation less about the "political situation" and more about the specific levers of power that could be pulled to prevent Allende from taking office.

What we have is a ghost of a conversation, filtered through two different agencies and a censorship pen that didn't even have the discipline to hide a name in the header. The cost of this curation was paid by the Chilean people, who were told their democratic process was being respected while the men in the room were simply confirming that the ground was ready for the blow.

Helms, Edwards and the September 18 Plan for Chile

"Discussion of the Chilean Political Situation"

Edwards and the Black Pen

Foggy Bottom's Filter

September 18, 1970

Sources