On March 8, 1968, Soviet Golf II-class submarine K-129 sank in the Pacific Ocean with 98 crew members, three nuclear ballistic missiles, and Soviet encryption equipment. Six years later, the CIA spent $800 million to build a custom deep-sea recovery vessel disguised as a commercial mining ship — and successfully raised a portion of the submarine from 16,500 feet below the surface. When the operation was exposed, the CIA created the 'Glomar response' — a legal formula that allows agencies to neither confirm nor deny the existence of records. The ship still exists. The cover story persists in maritime law.
On March 8, 1968, Soviet ballistic missile submarine K-129 sank in the North Pacific Ocean approximately 1,560 nautical miles northwest of Hawaii. All 98 crew members died. The Soviet Navy conducted an extensive search but never located the wreckage. The United States Navy knew exactly where it was.
The US Sound Surveillance System (SOSUS) — a network of underwater listening stations designed to track Soviet submarines — had detected an acoustic signature consistent with a catastrophic hull implosion. Naval intelligence analysts triangulated the event to a specific coordinate. Five months later, in August 1968, the specially equipped reconnaissance submarine USS Halibut located K-129's wreckage resting on the ocean floor at a depth of approximately 16,500 feet — more than three miles down.
USS Halibut made multiple passes over the debris field, deploying a specially designed camera sled called a "fish" that photographed the wreck in detail. The images revealed that K-129 had broken into sections. The conning tower (sail) remained largely intact but separated from the main hull. One missile tube door appeared open. The forward section of the pressure hull showed catastrophic damage.
The wreck represented an intelligence bonanza. K-129 carried three R-21 nuclear-tipped ballistic missiles, each with an 800-kiloton warhead. The submarine's cryptographic equipment, if recoverable, could provide access to Soviet naval communications. Missile guidance systems would reveal targeting capabilities and technological sophistication. Hull construction details would inform assessments of Soviet submarine vulnerabilities.
There was one problem: no technology existed to recover anything from three miles underwater.
In 1969, CIA Director Richard Helms commissioned a feasibility study. The question was not whether recovering K-129 would be valuable — that was obvious — but whether it was physically possible.
The study concluded that it was barely possible, extraordinarily expensive, and would require technologies that did not yet exist. Helms presented the proposal to President Richard Nixon and National Security Advisor Henry Kissinger. The estimated cost: $350 million, making it the most expensive covert operation in CIA history.
Nixon approved the mission. It received the codename "Project Azorian," though within the intelligence community it was often referred to as "Jennifer."
The CIA contracted Lockheed Missiles & Space Company as prime contractor. Lockheed's Ocean Systems Division had experience with deep-sea engineering from Navy bathyscaphe programs. The company would design and build the capture vehicle and heavy-lift system. Global Marine Inc., a commercial deep-sea drilling company, would provide operational expertise. Sun Shipbuilding & Drydock in Chester, Pennsylvania, would construct the recovery vessel itself.
The technical challenges were unprecedented. Engineers needed to design a ship with a center well large enough to accommodate a submarine section, capable of maintaining position in open ocean with precision measured in feet, and equipped with a lifting system rated for thousands of tons at a depth where water pressure exceeds 7,000 pounds per square inch.
The capture vehicle — nicknamed "Clementine" — required even more exotic engineering. The device would descend three miles on a string of custom pipe sections, each precisely engineered for strength-to-weight ratio. At the ocean floor, remotely operated hydraulic grippers would close around K-129's cylindrical hull. Then the entire assembly would lift the submarine section, pipe string, and capture vehicle back to the surface — a combined weight estimated at well over 2,000 tons.
And it all had to remain absolutely secret.
Constructing a 618-foot ship with capabilities unlike any commercial vessel in existence would attract attention. The CIA needed a cover story that would explain the Hughes Glomar Explorer's unusual design while discouraging detailed scrutiny.
The solution was Howard Hughes.
"Howard Hughes was perfect. He was reclusive, he had legitimate business interests in aerospace and defense, and he was known for eccentric projects that didn't always make immediate business sense. A deep-sea mining ship fit the pattern."
CIA official, quoted in David Sharp — The CIA's Greatest Covert Operation, 2012The CIA approached Hughes' Summa Corporation executives in 1970. Hughes himself, by that point deeply isolated and managing his businesses through a small circle of trusted executives, agreed to allow Summa Corporation to serve as the nominal owner and operator of the vessel. He received no payment; the arrangement strengthened his existing relationship with CIA proprietary companies and required minimal personal involvement.
The cover story positioned the Hughes Glomar Explorer as a commercial manganese nodule mining vessel. The story had advantages: deep-sea mining was a legitimate area of commercial research in the early 1970s, the technology was largely unproven (explaining experimental equipment), and the economics were speculative enough that questions about profitability could be deflected.
Summa Corporation filed patents for manganese nodule recovery systems. Trade publications received press releases about Hughes' entry into ocean mining. When Sun Shipbuilding began construction in 1971, the contract appeared in maritime industry databases as a commercial mining vessel.
Most of the construction workers, ship's crew, and even many Global Marine personnel believed the mining story. Security protocols compartmentalized access: crews worked in isolated areas, and the moon pool section was restricted. Only approximately 40 people knew the true mission.
The Hughes Glomar Explorer departed Long Beach, California, in June 1974. The ship proceeded to the target site in the North Pacific, approximately 1,560 miles northwest of Hawaii, and began station-keeping operations in early July.
The recovery operation proceeded in stages. First, the ship positioned itself over the target coordinates using a dynamic positioning system that fired computer-controlled thrusters to counteract wind and current, maintaining position within approximately 20 feet. Then the capture vehicle began its descent on the pipe string — a process that took several days as crews added pipe sections and lowered Clementine toward the ocean floor three miles below.
When Clementine reached the wreck, operators used sonar and cameras to position the capture vehicle over the submarine's hull. The hydraulic grippers closed around the target. The lift began.
For several days, the operation proceeded as planned. The pipe string bore the enormous weight as Clementine and the submarine section slowly ascended. Then, approximately two-thirds of the way to the surface, structural failure occurred.
Accounts differ on the exact mechanism. The most widely reported version, confirmed in later CIA declassifications, states that K-129's hull — weakened by 6 years on the ocean floor and the catastrophic damage from its sinking — fractured under stress. Approximately two-thirds of the grasped section broke away and fell back to the seafloor. Clementine retained only about 38 feet of the submarine's forward hull section.
The section that fell included the missile compartment — the primary intelligence target.
Despite the partial failure, the recovered section contained significant intelligence materials. The 38-foot forward hull section included:
CIA Director William Colby, who had taken office in September 1973, authorized preparations for a second mission to recover the remaining sections, particularly the missile compartment. The Hughes Glomar Explorer returned to port for repairs and modifications.
The second mission never occurred.
In February 1975, investigative journalist Jack Anderson received information from sources in the maritime industry who questioned the Hughes Glomar Explorer's stated mining purpose. Anderson contacted the CIA for comment. Director Colby personally met with Anderson, requesting that he suppress the story on national security grounds.
Anderson refused. His syndicated column on February 7, 1975, revealed that the ship had been built to recover a Soviet submarine. Some details were inaccurate — Anderson reported that the mission had fully succeeded, for instance — but the core story was public.
Colby contacted major newspapers, requesting that further details be withheld. The New York Times, Washington Post, and Los Angeles Times had been aware of the operation but agreed to delay publication based on CIA representations that exposure would compromise a sensitive intelligence operation. After Anderson's column appeared, the Los Angeles Times published a detailed two-part investigation on March 18-19, 1975, based on extensive independent reporting.
The Soviet government, now aware that the United States had recovered part of K-129, began monitoring the Hughes Glomar Explorer. Plans for a second recovery mission were cancelled. The operation was over.
But Project Azorian created one lasting legacy beyond the intelligence materials recovered from K-129's hull: the Glomar response.
In 1975, military researcher Phillippi filed a Freedom of Information Act request for documents related to the Hughes Glomar Explorer and Project Azorian. The CIA responded that it could "neither confirm nor deny" the existence of such records, arguing that even acknowledging whether records exist would reveal classified information about intelligence methods.
Phillippi sued. In 1976, the US Court of Appeals for the DC Circuit upheld the CIA's position in Phillippi v. CIA, 546 F.2d 1009. The court ruled that agencies could refuse to confirm or deny the existence of records when confirmation alone would compromise national security or protected privacy interests.
"The Glomar response has been invoked over 1,000 times since 1976 by federal agencies in FOIA litigation. What began as a doctrine protecting an extraordinarily sensitive intelligence operation has become a routine tool for avoiding transparency."
American Civil Liberties Union — FOIA litigation brief, 2018The Glomar response remains valid law. Federal agencies invoke it regularly, extending far beyond intelligence operations to law enforcement investigations, personnel records, and regulatory matters. Critics argue the doctrine has been overused to avoid legitimate transparency. The Supreme Court has never directly reviewed the Glomar response's constitutional validity.
The CIA began selective declassification of Project Azorian materials in 1992. A comprehensive historical account was released in 2010, coinciding with the 50-year anniversary of K-129's commissioning. These documents confirm key details while leaving some operational specifics classified.
The recovered section of K-129 provided significant intelligence value despite missing the missile compartment:
Nuclear Torpedoes: Two nuclear-armed torpedoes were recovered intact, providing detailed information about Soviet naval nuclear weapon design, arming mechanisms, and yield characteristics.
Cryptographic Materials: Codebooks and cipher equipment from the forward compartment gave Navy and NSA cryptanalysts access to Soviet naval communication protocols, though the intelligence value degraded over time as systems were updated.
Hull Construction: Analysis of hull metallurgy, welding techniques, and structural design revealed that Soviet submarine construction standards were lower than US intelligence had estimated, informing assessments of vulnerabilities.
Acoustic Signature Data: Propulsion and noise-reduction systems provided information useful for anti-submarine warfare and sonar development.
The CIA also recovered the remains of six Soviet sailors. In a gesture intended to demonstrate respect and potentially facilitate future cooperation, the Agency conducted a formal burial at sea with military honors, filming the ceremony. The footage was offered to the Soviet government in 1992 as part of post-Cold War intelligence sharing. Russia officially requested return of all recovered materials in 2003; the United States acknowledged the recovery and burial but declined to return materials citing continued classification.
After Project Azorian's exposure, the Hughes Glomar Explorer was mothballed. In the 1990s, the vessel was modified for legitimate deep-sea drilling operations, removing the specialized recovery equipment but retaining the moon pool and dynamic positioning systems. The ship worked under contract for various energy companies conducting offshore drilling exploration.
In 1997, the ship was acquired by Transocean, a major offshore drilling contractor, and renamed the GSF Explorer. It operated in this capacity until being retired from service in 2015. The vessel was sold for scrap and broken up in China.
Clementine, the capture vehicle, was reportedly destroyed after the mission. Lockheed engineers confirmed in later interviews that the device was dismantled and the components disposed of separately to prevent reconstruction of its design.
Project Azorian remains controversial among intelligence historians and former officials. Assessments vary significantly:
Supporters argue that the recovered materials provided valuable intelligence that could not have been obtained by other means, particularly regarding Soviet submarine construction standards and nuclear weapon design. They note that the intelligence justified the cost within the context of Cold War strategic competition, and that the operation demonstrated American technical capabilities that likely deterred future Soviet risk-taking.
Critics contend that $800 million for partial recovery of a six-year-old submarine wreck represented poor allocation of intelligence resources, particularly given that the primary target — the missile compartment — was not recovered. They argue that technical intelligence from the recovered materials had limited operational value given the time lag and ongoing Soviet technological development.
What is not disputed: Project Azorian represented one of the most ambitious covert engineering operations in history. It required developing technologies that advanced the state of the art in deep-sea operations. And it created a legal doctrine that continues to shape transparency and government secrecy debates five decades later.
The Hughes Glomar Explorer's story illustrates the infrastructure of Cold War intelligence operations: the coordination between intelligence agencies and private contractors, the use of commercial cover identities, the extraordinary costs justified by strategic competition, and the mechanisms by which operations remained secret — until they didn't.
The wreckage of K-129 remains on the Pacific Ocean floor. The section that fell back during the recovery operation sits where it landed, three miles down, in international waters. The Soviet Union never attempted to recover it. Neither has Russia. The technology required remains extraordinarily expensive. The intelligence value continues to diminish with each passing year.
But for one brief moment in 1974, the CIA built a ship unlike any other, deployed technologies that pushed engineering limits, and successfully raised a Soviet submarine from the deep. Most of it fell back into the darkness. What emerged was enough to confirm the operation's ambition — and to create a legal phrase that protects government secrets to this day.
The Glomar response: neither confirming nor denying. The perfect epitaph for Project Azorian itself.