The Cuba Report - factchecked
Is a war with Cuba in the offing?
The Cuba Report
The State Department released a Cuba Communism report on July 2026. When I read it, it seemed remarkably familiar to the “Dodgy Dossier” and Russian lies in the run up to the various Ukraine invasions. So I took a look, this is the result.
1. Factual errors and questionable specific claims
The “2,500 bombings” statistic is real but misframed. The report states the FBI counted ~2,500 bombings in an 18-month 1971–72 window and places it directly after Cuba’s ideological influence, implying left-wing (and by extension Cuban-inspired) authorship. The figure is genuine and traces to Bryan Burrough’s Days of Rage, but the underlying FBI count is of all domestic bombings in that period — which included right-wing, anti-Castro Cuban-exile, Puerto Rican nationalist, and a large mass of unattributed devices. Presenting it immediately downstream of Cuban ideological influence invites a causal read the raw number doesn’t support. (Rubio’s launch tweet does the same, adding “the overwhelming share...came from Left-Wing extremists” — a characterisation not in the FBI data itself.)
“Eighteen known SIGINT facilities in Cuba.” This is stated as flat fact with no citation. The CSIS “Hidden Reach” work the report itself relies on identified four active/under-construction sites (Wajay, Calabazar, El Salao, Bejucal). Where “eighteen” comes from — and the precise sub-attributions (”two belong to Russia,” “China actively runs three”) — is unsourced. The report leans on CSIS for credibility in one sentence, then produces a much larger, more specific number in the next without saying where it’s from.
Internal inconsistency on the Bejucal site. It’s described as both “the Soviet nuclear-warhead storage site... made famous during the Cuban Missile Crisis” and (via CSIS) a SIGINT facility. These are different function claims bundled together; the “nuclear-warhead storage” characterisation of Bejucal is contested in the historical literature and stated here as settled.
“Three years after taking power” / date compression on Castro’s Marxism. The report says Castro declared himself Marxist-Leninist “at the end of 1961, nearly three years after taking power.” He took power January 1959; end of 1961 is a little under three years — minor, but the “nearly three years” phrasing is doing rhetorical work to stress concealment.
Nayef Hawatmeh is misattributed. The report calls him “a founding member of the Popular Front for the Liberation of Palestine (PFLP) and the Democratic Front for the Liberation of Palestine (DFLP).” Hawatmeh founded the DFLP (he split from the PFLP to do so in 1969). Calling him a PFLP founding member is loose at best. It then says “Both popular fronts would later participate in the October 7, 2023, attacks” — the DFLP did claim participation, but folding these together compounds the sloppiness.
“Che Guevara arranged Cuba’s US law firm during a chess game.” Presented as fact, sourced only to “according to the Guild” — i.e. the target organisation’s own lore, repeated uncritically because it’s colourful and damning.
2. Poorly sourced or single-source claims
Heavy reliance on one defector. A large share of the espionage-scale claims — “90 percent of ICAP staff are affiliated with Cuban intelligence,” “less than 18 months to restore the Florida network,” DI involvement in “six of the initial eight” sister-city arrangements — rest on “Simmons” (Chris Simmons, a former DIA counterintelligence officer) or unnamed “former DI officers.” These are single-source, and Simmons is an advocate with a definite view. None of the 90% / 18-month / 6-of-8 figures carry independent corroboration in the text.
The Fox News investigation presented as established fact. The Singham network claims ($591 million, ~2,000 groups, $401m to US groups) are sourced to “a lengthy multi-part Fox News investigation” and “open-source analyses.” A partisan outlet’s investigation is treated as settled fact rather than as a contested allegation. Singham disputes these characterisations; the report doesn’t note that.
Anonymous officials carry structural weight. “According to one senior U.S. official,” “intelligence reports suggest,” “reports from 2025 allege” — the entire foreign-axis section (VIII) leans on unnamed sourcing for its most serious claims (Cuban drones from Iran, thousands of Cuban mercenaries in Ukraine, China/Russia “tripling” intelligence staff). “Reportedly” and “allege” are doing enormous load-bearing work.
The leaked “NNOC Rapid Response Plan” and the May union video call. Both are described as though authenticated (”clips leaked,” “began distributing”) without provenance, chain of custody, or any indication of how State verified them. For documents this central to the “fifth column” claim, that’s a notable gap.
3. Opinion and inference presented as finding
This is the largest category, and the pattern is consistent: real, documented facts are used as launch-points for sweeping interpretive claims stated in the declarative voice of an intelligence assessment.
The entire framing thesis is an interpretive claim. “Cuba is the ideological capital of the modern radical left” / “Castro was one of the fathers of the New Left” / Cuban Marxism “now enjoys something approaching ideological hegemony” — these are arguments, not findings. The report asserts causation (”shaped a disproportionate share of America’s most famous extremist movements”) where at most it can show contact, influence, and shared language.
The “civilisational” register. “A revolution against Western civilization itself,” “persuading the children of the West to turn against their own inheritance,” Cuba as “parasitic.” This is polemical editorialising, not analysis — and it recurs in the intro and conclusion, which are the least sourced sections of the whole document.
The explicit hedge that gives the game away. The conclusion states that George Floyd, Antifa, and campus activism “can be linked, in some way, shape, or form, to Cuban influence.” “In some way, shape, or form” is an admission that the connective tissue is associative, not evidentiary — yet it sits inside a sentence asserting Cuban authorship of major US political events.
DSA section openly concedes the absence of a link, then implies one anyway. To its slight credit the report says DSA “was not founded as a Cuban intelligence or influence front” and its commitment is ideological “not because the group is controlled by Cuban agents.” But it then spends pages treating DSA’s Cuba solidarity as functionally equivalent to control — the concession is rhetorical inoculation, not a genuine limit on the claim.
Guilt-by-association chains. Kathy Boudin → her son Chesa Boudin (former SF DA) is the clearest example: Chesa is named repeatedly purely by bloodline, contributing nothing but the suggestion that mainstream progressive prosecutors belong in a lineage with 1981 armed robbery. Similarly Zohran Mamdani, Ayanna Pressley, and Maxine Waters appear only via sympathetic statements about Shakur, positioned to imply a Cuban nexus that the text never actually establishes.
Motive attribution to a whole state. “Cuba’s fundamental objective remains... a hemispheric Marxist revolution that erases the United States from the earth” / “less a nation-state than an all-encompassing intelligence and subversion operation.” Ascribing a single, static, existential motive to an entire state across 67 years and multiple leadership generations is the report’s core inferential move, and it’s asserted, never demonstrated.
4. Selective framing and omissions
The acquittals and dropped charges are mentioned but rhetorically minimised. Angela Davis (acquitted), Huey Newton (charges dropped after hung juries), the report notes these in parentheses, then continues to file both under “black extremists,” letting the accusation outweigh the legal outcome it just acknowledged.
One-directional causation. The report never engages the obvious alternative reading, that US radical movements developed their own domestic grievances and sought out Cuba as a symbol, rather than being created by it. Influence is assumed to flow only Havana-outward. Allen Young’s 1970 quote (”Castro was one of the fathers of the New Left”) is an activist’s romantic self-description, deployed as though it were external corroboration of Cuban authorship.
The “AIPAC” aside. Flagging that Larudee’s group ironically shares the AIPAC acronym adds nothing analytical, it’s there for rhetorical effect, which in a document of this register is a tell about intent.
Conclusion: A Dossier in Search of a War?
The Cuba report shares a genuine rhetorical architecture with the Kremlin’s pre-invasion texts on Ukraine, collapsed history, denial of the target’s agency, civilisational framing, a hostage-population trope, real cases used as anecdotes, and escalating legal-administrative groundwork, but these structural parallels are not evidence of equivalent intent, and the report’s factual core is far better grounded than Moscow’s fabricated Donbas “genocide” claims.
The far closer and more defensible analogue is the Anglo-American Iraq dossier tradition of 2002–03: an assessment that took real intelligence and “portrayed a misleading picture of greater capability and greater certainty than the intelligence warranted,” written in a more assertive voice than intelligence documents traditionally use, and published to prepare public opinion, a warning that pretext-laying is a normal-statecraft pathology, not a uniquely Russian one.