The CIA secretly compiled an internal report, the "Family Jewels," cataloguing decades of its own illegal and improper activity

The record, stated plainly

The strength of this case is simply the strength of the primary source, and it is considerable. This is not a theory resting on anonymous tips; it is an internal government inventory of the government's own possibly illegal acts, later corroborated by two major investigations and then published by the agency that wrote it.

Consider what the file and the inquiries around it document. Domestic surveillance that the charter forbids: the CHAOS program built files on thousands of American citizens in the antiwar movement. Journalists targeted: the columnists Robert Allen and Paul Scott were wiretapped, and the investigative writer Jack Anderson and associates, including a young Brit Hume, were physically watched. Mail opened: under the HTLINGUAL operation, letters to and from the Soviet Union and China were intercepted and read, some screened at Kennedy airport.

There is more. A KGB defector, Yuri Nosenko, was held in isolation in conditions the file itself flagged as possibly violating US law. There were break-ins and warrantless entries tied to former employees. There were drug tests on unwitting subjects, connected to the wider MKULTRA research and to the 1953 death of Army scientist Frank Olson. And the Church Committee separately documented plots to assassinate foreign leaders, most fully Fidel Castro and Patrice Lumumba.

The accused agency compiled the list, Congress verified it, and the CIA itself eventually published it. On the core claim, that a secret file of real abuses exists, there is nothing left to prove.

That is the case at full strength, and it is unusually solid. Anyone who waved the whole thing away as paranoid fantasy would simply be wrong about the documented history.

Where care is still needed

Precisely because the file is real, it invites a second, looser move: treating it as a master key that unlocks every accusation ever made about the CIA. That step does not follow, and keeping it separate is the whole discipline of reading a document like this well.

Start with plots versus completed acts. The record establishes that the agency schemed to kill foreign leaders, in some Castro plots working through Mafia figures such as Johnny Rosselli and Sam Giancana. But the Church Committee also found that the CIA's own direct assassination attempts had not succeeded. Documented plotting is damning; it is not the same as a documented body count, and blurring the two claims more than the file supports.

Consider next the file's limits as evidence. It is a snapshot of what officers chose to report in 1973, not an exhaustive confession spanning all decades. Some relevant material, including many MKULTRA records, had already been destroyed before any accounting was possible. The 2007 release carried redactions. A bounded inventory with known gaps cannot logically certify things it never addressed.

And so the honest reading holds two things at once. The abuses the Family Jewels catalogues are real and serious, established beyond reasonable dispute. Yet the document does not thereby validate the long tail of unfounded claims later pinned to the agency. “The CIA did document real wrongdoing” is true. “Therefore any given accusation against the CIA is true” does not follow, and the file itself is the reason to keep those apart.

Why it came out at all

It is worth dwelling on how this record reached daylight, because the path matters to what it does and does not prove about secrecy.

The file was not volunteered. It was compiled defensively in 1973, in the shadow of Watergate, so the agency would know its own exposure. It stayed classified until Hersh's reporting forced public investigations, and even then the full compilation remained sealed. The National Security Archive's Freedom of Information Act request sat for roughly fifteen years before the release finally came in 2007, and that release still withheld portions.

That history genuinely supports a modest, well-earned conclusion: secret agencies do not disclose their worst conduct on their own timetable, and sustained outside pressure, from the press, from Congress, from FOIA litigants, is what pried this record open. That is a real lesson, and it is the reason the case remains cited.

What it does not license is the leap to bottomless suspicion, the idea that because this was hidden, everything imaginable is also hidden and equally true. The Family Jewels rewards skepticism aimed at documents and answered by documents. It does not reward the different habit of treating every non-disclosure as proof of a specific unproven crime.

Why the Family Jewels endures

Of all the intelligence stories of the era, this is the one that keeps being invoked, and it endures for reasons that are mostly to its credit.

It endures because it is verified. Most conspiracy claims ask you to trust a source over an official denial. This one is the official record, released by the agency itself, so it carries an authority that speculation never can. It is the case people point to when they want to show that institutional wrongdoing is not always imaginary.

It endures because it confirmed a prior. A large public already suspected that secret bodies overstep. When the proof arrived in the form of dated memos and named programs, it validated that suspicion in a way argument alone could not, and it made every later revelation feel more plausible.

And it endures because it is a cautionary reference point. The abuses drove the 1970s reforms, the standing oversight committees, and the framework meant to keep intelligence work inside the law. Whether those safeguards truly changed behavior, or mainly changed what gets written down, is exactly the sort of open question that keeps a fifty-year-old file in the present-tense conversation.