William Donovan and the 1963 Plan to Fracture Castro's Cabinet

By Harlan Pryce · September 18, 2026

A secret trip to Havana and a meeting in Palm Beach show a president abandoning exile assets to pursue a palace coup kept hidden from the diplomatic corps.

The cost of a failed intelligence asset is usually measured in burned networks and wasted years, but in the spring of 1963, the price was the strategic pivot of an entire administration. When the bridge to the Cuban exiles collapsed, the president stopped looking for an army to invade the island and started looking for a knife to twist inside the cabinet of the man he intended to destroy.

What survives here is a curated archival description rather than the record's own pages. The record consists of a summary of a meeting in Palm Beach, Florida, on April 15, 1963, and a subsequent memorandum regarding negotiations with the Cuban leadership.

The Failure of the Exile Bridge

By April 1963, the administration's reliance on the Cuban exile community had reached a point of systemic failure. The record captures a president who was no longer merely skeptical of his sources but openly frustrated by them. Specifically, the record states that the "President raised the question of Miro Cardona, stated that Cardona had misrepresented the facts."

José Miró Cardona had been a man of immense initial promise, serving as Cuba's Prime Minister for six weeks in early 1959 before his sudden resignation. He was the primary link between the White House and the anti-Castro elements, the man who could theoretically deliver the internal unrest the United States required. However, the connection between the president and Miró Cardona had turned toxic. The record reveals a leader who realized he was being fed a fantasy by a man desperate to maintain his relevance in the exile hierarchy.

This breakdown is the catalyst for everything that follows. The public record shows the president maintained a strategic objective to remove Fidel Castro, the leader of the Cuban revolution, from power. But the record of the Palm Beach meeting proves that the president had lost faith in the viability of the exile-led insurgencies. He was not abandoning the goal; he was abandoning the method. The pattern suggests that Miró Cardona’s "misrepresentation of facts" was the final proof that the exile networks were too compromised, too fragmented, or too dishonest to serve as the spearhead of a regime change operation.

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

The Donovan Backchannel

With the exile bridge burned, the administration pivoted to a different kind of asset: the professional clandestine operator. Enter William Donovan, the former head of the Office of Strategic Services (OSS) and a seasoned diplomatic negotiator. Donovan did not operate through the official channels of the State Department; he operated as a ghost.

Donovan traveled to Cuba to meet directly with Castro. According to the record, Donovan reported that Castro "knew that relations with the United States are necessary and Castro wanted these developed." This presentation of a pragmatic, isolated leader is a classic piece of tradecraft. Donovan's report continues, noting that there are "certain Cuban Government officials, communists, who are strongly opposed" to such relations. He adds that these officials, while under surveillance and lacking a great following, could plunge Cuba into "chaos" if they rebelled.

The desk's reading is that the "moderate Castro" narrative was a tactical framing used by Donovan to make a backchannel deal palatable to the hawks in Washington. By presenting Castro as a prisoner of his own hardliners, Donovan was not describing a diplomatic opportunity; he was identifying a fault line. If Castro was the pragmatist and the "Cuban Government officials"—the one-party Marxist-Leninist socialist republic's true believers—were the obstacle, then the solution was no longer an external invasion, but the amplification of that internal friction.

A Preference for Internal Sabotage

The Palm Beach meeting between the president and John McCone, the CIA director from 1961 to 1965, confirms this shift in appetite. The record is explicit: the president showed "skepticism about U.S.-based sabotage operations." He questioned "whether active sabotage was good unless it was of a type that could only come from within Cuba."

This is a critical distinction in tradecraft. U.S.-based sabotage—the kind involving exile raids, smuggled explosives, and CIA-backed paramilitaries—is loud, clumsy, and easily attributed to Washington. It creates a martyr of the target and justifies the target's crackdown on internal dissent. Sabotage that "could only come from within Cuba," however, is the hallmark of a palace coup. It is the work of the disgruntled colonel, the sidelined minister, and the purged party official. It is an operation that uses the target's own machinery to dismantle the regime.

The connection here is the timing. The president's frustration with Miró Cardona and his sudden interest in internal-only sabotage happen in the same breath. The administration had realized that the only way to remove Castro without triggering a global catastrophe was to make the removal look like a Cuban decision. The "chaos" Donovan mentioned regarding the hardline communists was not a risk to be avoided—it was the objective. If the hardliners rebelled, or if the pragmatists purged the hardliners, the United States could step into the vacuum.

The Dual-Track Deception

There is a telling gap in the architecture of these files. We have a "Memorandum for the File"—a summary of a meeting—and we know from the public record that there exists a separate document: "Donovan Cuban Negotiations," marked SECRET - EYES ONLY.

This is not a coincidence of filing. The desk's reading is that the administration was running a dual-track policy. The official record, and even the summaries shared with a limited circle, masked a more aggressive internal operation. While the summary focuses on the "necessity" of relations and the frustration with a specific exile, the "Eyes Only" files likely contained the actual mechanics of the destabilization plan.

Furthermore, the absence of the State Department from these communications is a glaring omission. The principals in this drama are a president, a CIA director, and a former intelligence chief. Not a single diplomat is mentioned. The pattern suggests the State Department was intentionally excluded to maintain plausible deniability and prevent official diplomatic interference. The administration didn't want a diplomatic solution; they wanted a clandestine result that looked like a diplomatic accident.

If the shape of this file is what it appears to be, the Donovan mission was never about "developing relations." It was a reconnaissance mission designed to map the internal fractures of the Cuban government. Donovan was not there to build a bridge to Castro; he was there to find the wedge. The "moderate Castro" was the lure, and the "hardline communists" were the target. By framing the conflict as one between a reasonable leader and an unreasonable cabinet, the United States could justify supporting the "reasonable" side—provided that support included the tools for a surgical, internal strike.

Who paid for this pivot? The Cuban officials under surveillance and the exile assets like Miró Cardona, who were discarded the moment they ceased to be useful. But more importantly, the American public was sold a narrative of diplomatic tension while the actual work was the cold, patient cultivation of a palace coup. The record shows a presidency that had moved beyond the hope of a friendly Cuba and had settled for a fractured one.

William Donovan and the 1963 Plan to Fracture Castro's Cabinet

The Failure of the Exile Bridge

The Donovan Backchannel

A Preference for Internal Sabotage

The Dual-Track Deception

Sources