OPERATION MONGOOSE

THE CLAIM

That the United States ran a sustained, presidentially authorised campaign of sabotage, economic warfare and assassination against Cuba, and that the plots to kill Fidel Castro were real operations with real budgets rather than rogue improvisation.

THE THEORY, AS ITS PROPONENTS TELL IT

The documented version of this file is already the strong version, which is unusual in this archive. Proponents simply argue that the record, once read, should change how every later allegation of covert action is weighed.

The programme was authorised at the top. Kennedy approved it in November 1961 after the Bay of Pigs, Edward Lansdale wrote the operational plan, and Task Force W ran it out of Miami with several hundred staff, its own navy of small boats, and a budget in the tens of millions. Sabotage of sugar and rail, contamination of exports, and propaganda broadcasts were all explicit objectives, and all were carried out.

The assassination plots are the part that reads as invention and is not. The Church Committee found at least eight, and the methods included a poisoned diving suit, an exploding seashell placed where Castro dived, cigars treated with botulinum, and a poison pen. The Agency contracted the Mafia, specifically Johnny Roselli and Sam Giancana, to do the killing, which is documented in the CIA's own Inspector General report of 1967.

From that starting point the wider argument is short. If an intelligence service will do this, write it down, and then keep it secret for fourteen years, the standard objection to conspiracy claims, that somebody would have talked, has already failed once at scale. That is the argument, and the documents are the evidence for it.

WHAT IS KNOWN

The programme is confirmed in the government's own record. The Church Committee's 1975 interim report, Alleged Assassination Plots Involving Foreign Leaders, laid out the plots against Castro alongside those against Patrice Lumumba and Rafael Trujillo. The CIA's internal Family Jewels review, compiled in 1973 and released in full in 2007, describes the Mafia contacts directly.

Operationally, Mongoose was large and largely unsuccessful. It did not remove Castro, it did not trigger the internal uprising Lansdale's plan assumed, and the Cuban Missile Crisis in October 1962 effectively ended the most aggressive phase. Sabotage operations continued into 1963 under other arrangements.

The Agency's own Inspector General concluded in 1967 that the plots were real, poorly run, and known to senior officials, while leaving the question of explicit presidential authorisation for the killing itself carefully unresolved. That ambiguity, deliberate or not, is the seam every later argument runs along.

EVIDENCE FOR

• The Church Committee interim report, a public congressional document, describing the plots in detail and naming the methods.
• The CIA's Family Jewels, the Agency's own compilation of activities it considered legally questionable, which includes the Mafia contract.
• Task Force W's size, budget and Miami station are documented in released records, as is Lansdale's operational plan with its month-by-month schedule.
• The 1967 Inspector General report, written for internal consumption and later released, which confirms the plots from inside the institution that ran them.

EVIDENCE AGAINST (THE WIDER READINGS)

• Mongoose does not establish the claims frequently attached to it. No released document connects the programme to the killing of President Kennedy, and the House Select Committee's interest in the Cuban operations produced suspicion rather than evidence. Compare OSA-022, where the same pattern holds: a great deal of new detail, no conspirators.
• The Agency's failure rate here argues against the omnicompetent version of the CIA that the wider theories require. Mongoose spent years and millions and achieved almost nothing.
• Assassination as a tool was formally prohibited by executive order in 1976, and no comparable programme has surfaced in the releases since, which is at least weak evidence that this was a period rather than a permanent condition.

ASSESSMENT

Confirmed, and worth reading precisely because it is confirmed. Everything in the operational core of this file, the sabotage, the propaganda, the Mafia contract, the poisoned diving suit, was denied, then rumoured, then documented. It sits in this cabinet as a calibration point: when someone says a government would never do that, this is the file to hand them. What it does not do is carry the weight later theories put on it. Mongoose proves the appetite and the machinery existed. It does not prove any particular thing they are accused of doing with it, and the distinction between those two claims is the whole discipline of this archive.

WHAT WOULD CHANGE THE GRADE

• CONFIRMED and needing nothing further: the Church Committee documented the plots in a public report, and the CIA's own Family Jewels and 1967 Inspector General review confirm them from inside the institution that ran them.
• The extension, that Mongoose or its Cuban networks connect to the killing of President Kennedy, would require a document, a participant account, or a financial link. Sixty years of releases, including the 2017 to 2023 batches, have produced none.
• Still sealed: portions of the Cuban operations record remain redacted, which is the only place an executed version of the wilder claims could still be sitting.

FURTHER READING

• Cuban Project (Operation Mongoose) · Wikipedia
• Church Committee · the 1975 interim report on assassination plots
• The Family Jewels · the CIA's own 1973 compilation, released 2007