Project Sundial: America’s secret plan to destroy the world in the 1950s

By 1952, American intelligence agencies had begun to warn that the Soviet Union might develop the capacity to carry out a successful first strike against the continental United States.

Military planners feared that such an attack could cut off national command centres and disable early warning systems before retaliation became possible.

As a result, Strategic Air Command and other analysts began to theorise backup plans that would, in theory, eliminate any advantage a Soviet first strike could achieve.

Within this strategic context, some proposals imagined a system that could, at least in principle, detect a full-scale attack and respond without authorisation from surviving human leadership.

Under the direction of General Curtis LeMay, military planners explored frameworks that required no presidential input or continuity of government and left no requirement for communication confirmation.

LeMay had previously directed saturation bombing campaigns over Japan, such as the firebombing of Tokyo in March 1945, and he consistently promoted strategies that were largely built on the assumption that overwhelming destruction ensured victory.

He once stated bluntly, "If you kill enough of them, they stop fighting," a logic that influenced many Cold War-era strategies.

Some early drafts of these ideas circulated among high-clearance military personnel and defence analysts who worked at the RAND Corporation.

While RAND’s involvement brought detailed modelling and systems planning, it also introduced early moral questions.

Analysts questioned whether any deterrent value could, even in theory, justify a plan that risked the end of the human species.

Still, some senior officials moved forward with general planning for second-strike capabilities that operated independently of leadership survival.

RAND figures such as Herman Kahn warned against automatic retaliation models, and they argued instead for more flexible strategies that later helped define the concept of Mutually Assured Destruction (MAD).

Under these proposed models, an automatic system would be expected to activate upon confirmation of several disastrous conditions, such as the destruction of command centres and verified thermonuclear detonations on American territory, combined with sustained failure of emergency communication systems.

Once those conditions aligned, the system would initiate a full-spectrum launch from all available nuclear platforms. 

At the core of this mechanism, a theorised “Red Line” protocol would act as a trigger system, which relied on a combination of radar inputs and seismic data.

It also drew on satellite telemetry and loss-of-signal detections from key defence outposts.

Inputs from the developing Ballistic Missile Early Warning System (BMEWS) were also considered for integration, expanding detection capabilities across the Arctic and Atlantic approaches.

If these inputs met the designated thresholds, the system was designed to bypass human decision-making and to issue immediate launch instructions to intercontinental ballistic missiles and nuclear bombers in flight, as well as submarine-based platforms, such as submarines equipped with Atlas and Titan-class delivery systems.

Significantly, some target lists in planning exercises apparently went well past the boundaries of direct Cold War adversaries.

Alongside primary targets such as Moscow and Leningrad in the Soviet Union and Beijing and Pyongyang in Asia, several theoretical models included strikes against major cities in non-aligned states such as India and Sweden, with Switzerland included in some drafts.

Defence planners justified this by asserting that, in their view, no alternative international order should rise in the absence of American power.

They argued that deterrence would only succeed if no future political authority could benefit from global collapse.

These plans, at least on paper, made no allowance for rescue operations, civilian evacuation, or post-conflict survival.

Instead, they framed global annihilation as the only acceptable consequence of a successful enemy first strike.

According to internal briefings shared between Strategic Air Command and the Department of Defense, such systems were intended, in the planners' view, to ensure that any attack on the United States would cost the aggressor everything, and even the future itself.

Between 1954 and 1957, RAND and Air Force War College staff ran multiple simulations to test automatic retaliation timelines and system reliability, together with the performance of fail-safe mechanisms.

Although no official reference to "Project Sundial" or a program called "Eclipse Trials" appears in declassified material, analysts did conduct hypothetical modelling of total communications failure and automatic nuclear response scenarios. 

Early tests used analog computing systems and punch-card data input to model potential failure points.

In most simulations, once a system confirmed loss of command and verified attack indicators, it could, at least in these models, issue full launch orders within nine minutes.

Given that Soviet ICBMs of the era had limited range and flight time, such speed was considered by many planners sufficient to guarantee retaliation even after disastrous losses.

At the same time, designers tried to introduce backup systems into each link of the chain.

Airborne command aircraft such as the EC-135 “Looking Glass” were given override capacity, but their destruction would not prevent automatic execution.

Instead, they acted as supplemental confirmation platforms. 

To prevent accidental activation, planners attempted to construct multiple layers of signal verification across independent detection sources.

However, they deliberately programmed the logic to favour deadly certainty over restraint.

If uncertainty in the signals existed, the system was to trigger rather than hesitate.

Defence analysts used the term “fail-deadly” to describe this logic, and this term meant that failure to verify peace would result in war.

Within RAND and among a few senior civilian officials, moral opposition to these automatic retaliation proposals grew by the mid-1950s.

Internal memoranda show that strategist Paul Baran described such ideas as “an irreversible suicide mechanism” that confused deterrence with self-destruction.

Baran had later carried out research that helped pioneer packet-switching systems that laid the groundwork for the internet, and he believed that sensible deterrence must retain flexibility.

Other strategists included Albert Wohlstetter, and they argued that uncertainty in response timing offered a stronger form of deterrence, and that removing decision-making from democratic control carried unacceptable risks.

Nevertheless, Strategic Air Command largely remained committed to the underlying logic.

LeMay’s senior staff defended automatic response doctrines on the basis that any gap in retaliatory capacity might encourage a Soviet first strike.

Defence officials such as Donald Quarles insisted that peace could only survive if adversaries knew that even a successful attack would offer no gain or advantage and no continuation.

Ethical concerns raised in these debates foreshadowed later theoretical discussions, such as the later idea of a nuclear winter that would render the planet uninhabitable even for neutral states.

At no point did any known automatic retaliation plan receive civilian ethical review.

Relevant documents remained restricted under Special Access Programs, and neither the White House nor Congress appears to have had access to full operational parameters.

Public debates about deterrence strategy during the Eisenhower years largely focused on visible weapons systems, while these theories stayed hidden behind high-clearance tight secrecy.

As nuclear strategy shifted during the early 1960s, the Kennedy administration began to reconsider fixed-response policies.

Robert McNamara’s advocacy for flexible response models led to a series of planning reforms that reduced the importance of total war scenarios.

During this period, ideas about automatic retaliation such as those theorised earlier gradually lost their use in real war plans, and by 1963, their logic no longer aligned with official American doctrine.

New doctrines, such as McNamara’s "No Cities" proposal, openly rejected the principle of targeting civilian centres as a strategic baseline.

Still, some elements of these earlier planning models persisted. Their automation logic influenced later strategic documents and helped guide the design of fail-safe mechanisms embedded in the Single Integrated Operational Plan (SIOP).

Additionally, Soviet planners developed a confirmed system known in the West as “Dead Hand,” which bore a similar idea to theorised American Red Line protocols.

Events such as the 1995 Norwegian rocket incident appear to have triggered temporary panic among Russian command due to its resemblance to a U.S. first strike and demonstrated how remnants of the fail-deadly mindset lingered long after these proposals fell out of favour. 

After decades of secrecy, documents related to automatic retaliation theory had appeared in Freedom of Information Act releases in the 1990s.

Declassified technical notes and systems outlines, which now sit in the National Security Archive, show how seriously many officials in the United States once treated total destruction as a strategic necessity.

Within those documents, the logic still appears stark: if the United States ceased to exist, then no world should continue without it.

Stanley Kubrick’s film Dr. Strangelove (1964) satirised automatic doomsday logic and echoed the very ideas that Cold War strategists debated in secret.