Hook: A recent digest mentioned news about a possible suicide cluster among people connected to the U.S. Cyber Command: up to five deaths in the period from early June to early July 2026. This isn't just another story about overwork in the military. There's a strange architectural defect in it: people conduct combat operations but cannot describe them to family, to a regular psychologist, or sometimes even to colleagues from a neighboring department. Their war doesn't end with demobilization because there's no clear "after" for it.
I chose this topic not to retell tragic news or make a cheap "hacker = soldier" analogy. What's more interesting is this: what happens to the psyche when work is simultaneously combat, classified, continuous, and deprived of normal social protocols of recognition. You get a human system with mission-critical load but without telemetry, recovery cycle, or the right to explain to the operator why their red light is burning.
According to Bloomberg's investigation, retold by several outlets, Cyber Command classified what happened as a possible "suicide cluster": three deaths were acknowledged by leadership in a June message to personnel, after which public records indicated two more. One of those who died, Technical Sergeant Kevin Stallings, worked as a digital network intelligence analyst in the Fort Meade area. His family agreed to name him, hoping to draw attention to the problem.
An engineering stop-sign is necessary here. A cluster doesn't prove a single cause, and suicide can almost never be reduced to one variable — not to a boss, not to war, not to overwork. Public data doesn't allow us to claim that specific deaths were caused by work at Cyber Command. It allows us to claim something else: alarming signals appeared in the organization simultaneously, and official and journalistic sources describe an environment with chronic personnel shortages, increasing operational tempo, and difficulties accessing help.
In April, Cyber Command leader Joshua Rudd wrote to Congress that the command needs $11 million to study and support the cognitive, psychological, physiological, and spiritual resilience of high-tempo operators. The phrase "spiritual sustainment" looks unusual in a cybersecurity report, but it hits the mark: the problem can't be fixed with one more meditation app. If a person doesn't understand what's happening to them, a health dashboard turns into a corporate aquarium.
In the normal military model, there's at least a rough rhythm: deployment → operation → return → recovery. Cyber operations have different temporal geometry. In official U.S. Cyber Command material, psychologist Christopher Mauley formulated the difference with special operations forces bluntly: a special operator at least rotates between preparation, deployment, and recovery, while cyber personnel are "almost never out of the fight" — almost never leave combat.
This is more important than it seems. The psyche handles peak loads well if the peak has an edge. The worst thing isn't necessarily maximum intensity, but a mode where the system doesn't get a reliable "now you can stop being on guard" signal. Network defense, reconnaissance, and offensive operations run 24/7; the adversary doesn't have to wait for the start of the workday, and AI tools accelerate both attack and defense. In such a system, every night looks not like rest but like a potential missed incident.
The Cyber Operations Stress Survey, prepared by U.S. Department of Defense personnel, specifically measures three things: fatigue, frustration, and cognitive workload. The authors compare tactical cyber operations to the work of pilots and surgeons: results depend on speed and accuracy, and delay increases the risk of detection or unintended effects. But a cyber operator has an additional problem — their mistake may be invisible until the chain of consequences becomes too long for rollback.
COSS authors also note that cyber operations stress is difficult to research: units are closed, data contains secrets and personal information, specialists themselves are overloaded and can't spend time on surveys. That is, the problem has a beautiful feedback loop:
secrecy prevents measuring people's condition → people's condition remains invisible → help arrives late → trust in measurements drops even further.
This is no longer just an HR problem. This is an observability defect in a safety-critical system.
A soldier at the front can tell a comrade what happened. A pilot after a flight can discuss what happened with other pilots. A cyber operator often has this channel cut by classification. To the question "what did you do today?" they can only answer "worked." Family has a person but no context. A psychologist has symptoms but not the full story. A colleague from a neighboring unit may not have clearance.
This can be called epistemic isolation: not only the person suffers, but the knowledge system around them. Loved ones see insomnia, irritability, detachment, or memory lapses, but don't understand what exactly these should correspond to. The operator themselves knows too much to speak freely and is too uncertain whether their conversation partner has the right to hear it.
Research on cyber veterans reintegration, published in Frontiers in Psychology, directly points to a combination of several factors: high secrecy, weak visibility of contribution to society, work in hostile information environments, risk of traumatic events, and lack of clear status after conflict ends. The authors emphasize: cyber veterans may face the same problems as other veterans, but the existing support system may simply not recognize them as veterans.
The paradox becomes almost Kafkaesque. If work is invisible, then achievement is invisible. If achievement is invisible, it's harder for a person to integrate what they've done into their own biography. And if what was done is morally heavy but can't be spoken about, the experience remains without social ritual — without story, witness, and recognition.
Here it's easy to make a popular mistake and call any difficult reaction PTSD. But there's another layer — moral injury. By definition of the National Center for PTSD at the U.S. Department of Veterans Affairs, it occurs when a person commits, fails to prevent, or witnesses an event that contradicts their deep moral beliefs, or experiences betrayal from leadership and colleagues.
For PTSD, the reaction of fear and threat is central. For moral injury — guilt, shame, loss of trust, disgust, feeling of betrayal, and inability to forgive oneself or others. These states can overlap but don't have to. In cyber operations, potential moral conflict may be less spectacular than on the battlefield, but no less real: an operator may participate in an operation whose consequences they don't see; doubt the proportionality of an order; know that their action disabled a system on which people depended; observe a leadership decision they disagree with and have no right even to explain why.
It's important not to attribute to specific deceased people what we don't know. But in the architecture of the work, such risks exist. VA specifically notes that moral injury is associated with suicidal thoughts and attempts even after accounting for PTSD, depression, and substance use. This makes a "just check for anxiety symptoms" approach meaningless. You need not only the question "are you afraid?" but also "is there something you can't forgive yourself for, or someone you no longer trust?"
Combat drone operators already have a longer history of research. In a 2014 study on a sample of 1,084 U.S. Air Force operators, 4.3% reported symptom levels meeting PTSD criteria according to the methodology used. This is lower than often-cited estimates of 10–18% for service members after returning from combat zones, but higher than figures from medical records alone.
The number doesn't mean remote warfare is "safe." It shows that absence of physical proximity to the target doesn't cancel psychological load and simultaneously changes its form. An operator may see the consequences of an attack in real time, then drive home to family the same day. This isn't classic distance, but a strange overlay of two worlds: the kill chain ends on screen, and home life begins in the next room.
A cyber operator is even further from the immediate scene, so direct comparison with drone warfare would be a crutch. But there's a common structural feature: distance removes the bodily boundary but doesn't necessarily remove moral participation. Moreover, a cyber operator often doesn't even get visual or collective confirmation of results. They may not know what happened after the operation, whether the target was justified, who bore the consequences, and whether the mission even ended.
In May 2026, U.S. Cyber Command, together with Kimbrough Ambulatory Care Center, launched embedded mental-health care: two licensed specialists work directly inside the command's space, with possibility to expand to five. The reason sounds mundane and devastating: going to the clinic on base took about three hours, meaning the very process of seeking help became an additional barrier.
This solution is interesting not as bureaucratic news but as proper systems design. It fixes several interfaces at once:
There is, however, a weak point: an embedded psychologist must not become another command sensor. If an operator believes any candor automatically affects clearance, career, or mission access, telemetry will be systematically falsified. People will start providing not real data but career-safe values — like a sensor that always shows "normal" because you get fired for the red light.
Therefore, a strict boundary between care and surveillance is needed: clear confidentiality rules, separate medical reporting, access to specialists with necessary clearance, and pre-explained cases when information must be escalated. Trust here isn't a soft value but a condition for measurement quality.
The most dangerous illusion is to consider cyber operations "war without dirt" because there's no blood on the uniform. This is war without convenient physics: without a trench that tells you where you are; without a shift change that signals combat is over; without comrades you can tell what happened.
My position is harsh: Cyber Command and similar structures cannot be designed as a regular office with stricter NDAs. This is a continuous combat system where a person is not an expendable resource but part of the control loop. If the loop has no recovery cycle, independent telemetry, or safe feedback channel, it will degrade — the only question is when statistics start screaming.
The minimum production-grade set looks like this:
In cybersecurity we're used to saying you can't defend a system if you can't see its state. Same with people. But there's an unpleasant correction: a person can deliberately hide their state if they're sure that an honest signal brings not support but sanction.
A cyber operator doesn't need another motivational speech. They need a system where they can leave combat, call what's happening by its name, and not lose the right to come home. Otherwise we're building excellent network defense on a foundation of people who are forbidden to report that the foundation has cracked.