Project ARTICHOKE: The CIA Interrogation Program That Became MK-Ultra.

What ARTICHOKE was, in a paragraph.

Project ARTICHOKE was a CIA program of human interrogation research that ran formally under that name from August 1951 until April 1953, having begun in April 1950 as Project BLUEBIRD. Its institutional home was the agency's Office of Security; its scientific direction came from the Office of Scientific Intelligence and a panel of outside medical and psychological consultants. The program's stated objective, in the language of its own authorizing memos, was to develop "special interrogation techniques" against captured foreign agents and defectors, using combinations of drugs (notably sodium pentothal, sodium amytal, mescaline, marijuana derivatives, and from 1951 onward LSD-25), hypnosis, sensory disruption, and physical coercion, including in some sessions electroconvulsive treatment. Two specific operational questions drove the research. The first — the "defensive" question — was whether American personnel captured abroad could be made to disclose classified information against their will, and what countermeasures might be possible. The second — the question the agency's later defenders have most strenuously denied was ever seriously pursued and the documentary record most clearly shows was pursued — was whether a subject could be induced under hypnosis or drugs to commit an act, including a violent one, against their conscious will and without their later memory. ARTICHOKE generated extensive internal records during its life; in April 1953 its research lines were transferred to the newly authorized Project MKULTRA under the Technical Services Staff, and ARTICHOKE itself was administratively dissolved. The 1973 destruction of MKULTRA records on the order of Director Richard Helms swept up most of ARTICHOKE's later operational material with it. What survives is the program's first two years — substantial but incomplete — in the form of memoranda, status reports, contract files, and consultant correspondence recovered through Freedom of Information litigation from 1977 onward.

The documented record.

BLUEBIRD: April 1950 — August 1951

The program now known as ARTICHOKE was originally authorized on Verified April 20, 1950, in a memorandum from CIA Director Roscoe H. Hillenkoetter establishing "Project BLUEBIRD" under the Office of Security [1]. The name change to ARTICHOKE was made on August 20, 1951, shortly after CIA Director Walter Bedell Smith took office. The change appears in the surviving record to have been driven by routine compartmentation rather than a substantive shift; the personnel, the consultants, and the lines of research carried forward intact [2].

The earliest BLUEBIRD field activity that survives in the record is a series of Verified interrogations conducted in Tokyo in July 1950 on four suspected double agents, in which the subjects were administered combinations of sodium amytal, sodium pentothal, and Benzedrine. Internal reports characterized the results as inconclusive: subjects became disoriented and verbose but the interrogators could not distinguish reliable disclosures from fabrications [3]. A second series in Frankfurt later in 1950 produced similar findings.

The Office of Security and the Scientific Panel

BLUEBIRD and then ARTICHOKE were administratively housed in the CIA's Office of Security, then directed by Sheffield Edwards. Operational responsibility within the Office sat with the Security Research Staff under Morse Allen, whose name appears as author or recipient on a large share of the surviving memoranda. The scientific oversight of the program came from a separate body, the agency's Office of Scientific Intelligence, and from a rotating panel of outside consultants, including the psychiatrist Harold Wolff of Cornell, the pharmacologist Harold Abramson of Mount Sinai, and at various points anesthesiologists, hypnotists, and toxicologists drawn from Boston-area and Washington-area hospitals [2][4]. Verified

Hypnosis research

Hypnosis was a central line of inquiry. Verified A January 1952 ARTICHOKE memorandum, authored by Morse Allen and partially released in 1977 with redactions and again with reduced redactions in 2018, describes an experiment in which a young female subject — identified only by initials — was induced under hypnosis to "fire" what she was told was a loaded pistol at another subject (the pistol was unloaded). The memorandum records that she did so, and that on being awakened she had no memory of the act and "refused to believe she had pulled the trigger." Morse Allen's own report concludes: "This, of course, was a clear-cut example of a successfully induced post-hypnotic suggestion against the will of the subject" [5].

Whether this experiment generalizes — whether a hypnotic subject in operational conditions, knowing the firearm to be loaded and the target real, would carry out the act — is left open in the surviving record. Internal ARTICHOKE memoranda from later in 1952 and 1953 record both optimistic ("the technique appears feasible") and pessimistic ("the safeguards inherent in the human personality cannot be reliably overridden") assessments. Both views were held simultaneously inside the program. Claimed

Drug research

ARTICHOKE conducted or contracted out drug testing on a substantial scale during its existence. Verified Sodium pentothal and sodium amytal — the so-called "truth serums" then in use medically — were tested in interrogation conditions on both volunteer and unwitting subjects. Mescaline and tetrahydrocannabinol derivatives were tested as disinhibitors. LSD-25, which had been first identified to the CIA's interest in early 1951 through a Sandoz Laboratories connection in Switzerland, entered ARTICHOKE's drug inventory in late 1951 and was the focus of substantial work through 1952 and into 1953 [4][6]. The program also evaluated heroin and cocaine in similar contexts.

Subject populations for these drug tests included U.S. Army volunteers under contract at the Edgewood Arsenal in Maryland (later expanded under separate Army programs that became the Edgewood Arsenal experiments), prisoners (notably at the U.S. Penitentiary in Atlanta and the federal narcotics facility in Lexington, Kentucky), and individuals identified as suspected hostile agents in the field. The Lexington studies were conducted in cooperation with Dr. Harris Isbell, whose subject pool was drawn from incarcerated addicts who reportedly were paid in the drug to which they were addicted [4][7]. Verified

The Pohle and other field operations

ARTICHOKE's operational arm conducted at least a dozen documented field interrogations of foreign nationals between 1951 and 1953, several of which survive in summary form in the record. Verified The most discussed in the historiographic literature is the 1952 interrogation in Germany of a man identified in the partial release as "Pohle" — a suspected Soviet-recruited double agent — using a combination of barbiturates, sleep deprivation, and the threat of further drug administration. The case officer's after-action report describes the subject as having ultimately confessed details that were later partially corroborated by other intelligence, but cautions that confession-under-drugs cannot be assumed reliable [8].

Transfer to MKULTRA

Project MKULTRA was authorized by Director Allen Dulles on Verified April 13, 1953. Its formal scope — "research and development of chemical, biological, and radiological materials capable of employment in clandestine operations to control human behavior" — substantially overlapped ARTICHOKE's. Within weeks of MKULTRA's authorization, ARTICHOKE's surviving operational case files, scientific consultants, and the active drug-testing contracts were transferred to the new program under the Technical Services Staff. ARTICHOKE itself was not formally cancelled by a single memorandum; the program simply ceased generating new subprojects and its remaining operational activity was absorbed [1][9]. By the end of 1953, the Office of Security retained only a small residual interrogation-research function under the name; by 1955, the name appears in the agency's internal lexicon only as a historical reference.

The 1975 acknowledgment

The existence of BLUEBIRD/ARTICHOKE was first publicly acknowledged by the CIA in Verified 1975, in testimony to the Church Committee, alongside the broader disclosure of MKULTRA. The committee's final report described ARTICHOKE only briefly, identifying it as the institutional predecessor of MKULTRA and noting that its records were largely no longer extant. A more substantial picture emerged only after John Marks's 1977 FOIA recovery of MKULTRA financial records, several hundred pages of which referenced or summarized earlier ARTICHOKE material [10]. The most recent material release came in Verified 2017–2018, when the CIA released to the CIA Reading Room (cia.gov/readingroom) approximately 1,300 additional pages of ARTICHOKE-related correspondence under continuing FOIA litigation by The Black Vault and other researchers [11].

The official explanation.

The position the CIA has held consistently since 1975 is that ARTICHOKE was an interrogation-research program initiated in response to a credible Cold War threat — specifically, the suspected Soviet, Chinese, and North Korean use of chemical and behavioral methods on captured Western personnel. The 1949 show trial of Hungarian Cardinal Jozsef Mindszenty, whose public confession appeared to American observers to be drug-induced or otherwise coerced, is cited in the program's own founding memoranda as a motivating event. The Korean War period, with its later allegations of "brainwashed" American POWs, sharpened the perceived urgency. The program's defensive aim — understanding what techniques might be used against U.S. personnel and what countermeasures existed — is treated by the agency as the primary and legitimate purpose. Claimed

The agency's position on the offensive question — whether ARTICHOKE sought to develop techniques to induce a subject to act against their will — has been more careful. The 1977 Senate testimony of Director Stansfield Turner acknowledged that such research was undertaken but characterized it as a limited and ultimately unsuccessful line of inquiry. Internal CIA reviews from 1963 (the Earman Inspector General report) and from later periods reach a similar conclusion: that no operational capability of the kind described in the more dramatic public accounts was ever produced. The agency does not, in its official position, deny that the research was attempted; it denies that it succeeded [10][12].

The institutional position on the subjects' consent has been to acknowledge that some subjects of ARTICHOKE research were unwitting, particularly in the field-interrogation work, but to argue that the laboratory-setting drug tests were generally conducted on contracted volunteers and prison populations who had signed forms whose adequacy under modern standards the agency does not defend.

The unanswered questions.

The 1953 transfer records

The most significant single gap in the surviving record is the body of ARTICHOKE material that moved with the program into MKULTRA in spring 1953. Those files would, if they survived, document both the program's terminal state of knowledge and its outstanding research lines. They did not survive: the 1973 MKULTRA destruction order swept them up along with the rest of MKULTRA's operational material. The roughly 20,000 pages that survived because they had been misfiled in MKULTRA's financial records contain ARTICHOKE references but not operational reports. The cumulative result is that the period covered by surviving records ends, for most lines of research, in late 1952 or early 1953.

The hypnosis question

Whether the work documented in the January 1952 Morse Allen memorandum — the unloaded-pistol experiment — was extended to operational conditions, and what was learned, is unresolved on the public record. Disputed Researchers including John Marks, Alan Scheflin, and Edward Opton, working from the surviving fragmentary record and subsequent declassifications, have argued that the work was extended, that a small number of operational experiments were attempted, and that the program's terminal assessment was that reliable hypnotic control could not be produced [7][13]. The agency's position is that the work was attempted but did not yield an operational capability. The two positions are not as far apart as they sometimes appear in the popular literature, but the difference between "attempted and failed" and "attempted, partially succeeded, and then abandoned" remains substantively unresolved.

The total number of subjects

No comprehensive subject list exists in the public record. The surviving documents identify individuals only by initials or by case-codes. The Lexington-Atlanta drug studies, the Edgewood Arsenal contract testing, and the field interrogation series together produced subject populations whose total size has been estimated, by historians including Marks (1979) and H. P. Albarelli (2009), at "several hundred" through ARTICHOKE's existence [10][14]; that estimate is not derived from a counted roster but from extrapolation. Unverified

Specific deaths under ARTICHOKE

At least two deaths during ARTICHOKE-period drug testing are referenced in surviving documents without being detailed. A 1952 internal memorandum mentions a subject who died during a sleep-deprivation and drug administration sequence; the subject's name, location, and details are redacted. A separate reference describes a fatal cardiac event during a barbiturate-LSD test, with the same level of redaction. Whether either case has been independently corroborated outside the agency's own documents is, as of 2026, not publicly resolved [11][14]. Disputed

The Frank Olson question, in its ARTICHOKE light

Frank Olson's Disputed November 1953 death — covered in detail in our MK-Ultra file — technically occurred under MKULTRA, which had absorbed ARTICHOKE's drug-testing operations seven months earlier. The agency that administered the LSD to Olson at Deep Creek Lake was the team that had been doing the same work under ARTICHOKE's name through 1952 and 1953; only the program's letterhead had changed. Whether Olson's death is properly described as an MKULTRA event or as the last ARTICHOKE event is partly a matter of which name was on the budget code at the time [15]. The substantive record is the same regardless of the label.

Primary material.

The accessible primary record on Project ARTICHOKE is held at four locations:

• The CIA's FOIA Reading Room (cia.gov/readingroom) hosts the most current consolidated release of ARTICHOKE material, including the 1950 Hillenkoetter authorizing memo, the 1951 name-change memorandum, a substantial run of Morse Allen's program correspondence, and the 2017–2018 supplemental releases.
• The National Security Archive at George Washington University holds John Marks's collection of ARTICHOKE references recovered through his 1977 MKULTRA FOIA work.
• The National Archives (NARA) holds the Church Committee's final report and supporting documents, including the brief ARTICHOKE-specific testimony.
• The Black Vault (theblackvault.com), maintained by John Greenewald Jr., has hosted parallel batches of the 2017–2018 releases and corresponded with the CIA's FOIA office on the chain-of-custody documentation.

Critical individual documents include: the April 20, 1950 BLUEBIRD authorization memorandum; the August 20, 1951 name-change memorandum to ARTICHOKE; the January 1952 Morse Allen "post-hypnotic suggestion" memorandum (the unloaded-pistol experiment); the September 1952 ARTICHOKE Program Status Report; and the spring 1953 series of transfer memoranda moving ARTICHOKE's research lines into the newly authorized MKULTRA.

The sequence.

1. April 20, 1950 CIA Director Hillenkoetter signs the memorandum authorizing Project BLUEBIRD under the Office of Security.
2. July 1950 First documented BLUEBIRD field interrogations conducted in Tokyo on four suspected double agents.
3. February 1951 CIA first formally identifies LSD-25 as a substance of operational interest, through Sandoz Laboratories contact.
4. August 20, 1951 Project BLUEBIRD renamed ARTICHOKE under a memorandum from Director Walter Bedell Smith.
5. January 1952 Morse Allen prepares the "post-hypnotic suggestion" memorandum describing the unloaded-pistol experiment.
6. Spring–Fall 1952 Expanded drug testing program at the federal narcotics facility in Lexington, Kentucky, under Dr. Harris Isbell.
7. September 1952 ARTICHOKE Program Status Report summarizing program state and outstanding research lines.
8. April 13, 1953 Director Allen Dulles authorizes Project MKULTRA. ARTICHOKE's research lines, consultants, and active contracts transfer to the new program within weeks.
9. November 19, 1953 Frank Olson dosed with LSD at the Deep Creek Lake retreat, under what was nominally an MKULTRA-administered session but was operationally continuous with ARTICHOKE's drug work.
10. November 28, 1953 Olson falls to his death from the tenth floor of the Hotel Statler, New York.
11. 1973 CIA Director Helms orders destruction of MKULTRA records, which by then included the absorbed ARTICHOKE files.
12. 1975 Church Committee testimony first publicly identifies BLUEBIRD/ARTICHOKE as the institutional predecessor of MKULTRA.
13. 1977 John Marks's FOIA recovery of MKULTRA financial records produces several hundred pages of ARTICHOKE references and summaries.
14. 1979 John Marks publishes The Search for the "Manchurian Candidate", the first sustained public account.
15. 2017—2018 CIA releases approximately 1,300 additional pages of ARTICHOKE-related correspondence under continued FOIA litigation. These supplement, but do not replace, the operational record lost in 1973.

Cases on this archive that connect.

MK-Ultra (File 001) — the direct successor program that absorbed ARTICHOKE's research lines in spring 1953. The Frank Olson death sits at the institutional seam between the two programs.

COINTELPRO (File 009) — the contemporaneous FBI counterintelligence program. ARTICHOKE and COINTELPRO operated in adjacent but distinct domains: ARTICHOKE against foreign subjects, COINTELPRO against domestic political organizations.

More related files coming as the archive grows. Planned: Project BLUEBIRD as a standalone file, Operation Midnight Climax, the Edgewood Arsenal experiments, Project MKNAOMI, the Church Committee documents.

Full bibliography.

1. Memorandum from Director of Central Intelligence Roscoe H. Hillenkoetter, "Project BLUEBIRD," April 20, 1950. CIA FOIA Reading Room.
2. Memorandum on renaming of BLUEBIRD to ARTICHOKE, August 20, 1951. Office of Security records, partially released 1977 and 2017.
3. ARTICHOKE field interrogation summaries, Tokyo series, July 1950. Released 1977 and re-released with reduced redactions 2018.
4. Office of Scientific Intelligence consultant correspondence, 1951–1953, partially released 2017–2018.
5. Morse Allen, "Memorandum on Post-Hypnotic Suggestion in Experimental Settings," January 1952. Released to John Marks via FOIA, 1977; reduced-redaction release 2018.
6. Sandoz Laboratories correspondence with the CIA's Office of Scientific Intelligence, early 1951. References to the Hofmann synthesis of LSD-25 and the agency's commercial interest in supply.
7. Marks, John, The Search for the "Manchurian Candidate": The CIA and Mind Control, Times Books, 1979. The first comprehensive public account, drawing on the 1977 FOIA recovery.
8. "Pohle" case after-action report (subject's identifying detail redacted), 1952. Released 2017 in 1,300-page supplemental release.
9. Memorandum from Allen W. Dulles, "Project MKULTRA," April 13, 1953. CIA FOIA Reading Room.
10. Senate Select Committee to Study Governmental Operations with Respect to Intelligence Activities (Church Committee), Final Report, Book I, 1976. Coverage of ARTICHOKE at pages 385–395.
11. The Black Vault (Greenewald, John), ARTICHOKE 2017–2018 release archive. URL: theblackvault.com. Chain-of-custody documentation for the 1,300-page supplemental release.
12. CIA Office of Inspector General, "Report of Inspection of MKULTRA," John S. Earman, July 26, 1963. The earliest internal review reaching the program's predecessor work.
13. Scheflin, Alan W. and Opton, Edward M. Jr., The Mind Manipulators, Paddington Press, 1978.
14. Albarelli, H. P., A Terrible Mistake: The Murder of Frank Olson and the CIA's Secret Cold War Experiments, TrineDay, 2009.
15. Starrs, James E., "The Death of Frank Olson: Cause of Death," forensic report, 1994. Held in records of the Manhattan District Attorney's Office; relevant here for the ARTICHOKE-to-MKULTRA institutional continuity.