Every system that acts on the world needs a legible object. An intelligence apparatus needs a target it can name, locate, and strike. A regulatory framework needs a threat it can define, measure, and legislate against. A search index needs documents it can rank and cite with confidence.

When the real object does not fit the system’s grammar, the system does not stop. It fabricates an object from its own categories and operates on that with full confidence.

This is not malfunction. It is the system working exactly as designed.

The phantom target#

Iraq, 2003. The intelligence corridor needed weapons of mass destruction — a legible target that justified the machinery already in motion. The target was fabricated from the corridor’s own grammar: satellite photos annotated with certainty, defector testimony shaped to fit, dissent filtered out at every checkpoint. The output looked identical whether the object was real or invented. The corridor could not signal the difference, because the difference was invisible from inside.

Only materiality broke the spell. No weapons found. The phantom became visible, but only after the operation was complete.

Now look at the current war against Iran. As the conflict escalates, the language has drifted from defensive strikes toward regime change — a familiar grammar. But the Islamic Revolutionary Guard Corps is not a regime in any legible sense. It is a decentralized network of commanders, proxies, and institutional interests with no single head to remove. The corridor needs a “regime” because regime change is a grammar it knows how to operate. So it fabricates one, and the operation proceeds as if the object were real.

The pattern is not new. What is worth noticing is how reliably it repeats, and how little the previous failure teaches. Iraq’s phantom was eventually exposed by the ground. Iran’s phantom may never be, because a decentralized network can always be re-described as a regime that has not yet been properly decapitated.

The phantom guardian#

Here is where it gets worse. The corridor does not only fabricate the target. It also fabricates the guardian — the entity that defines what “safe,” “protected,” and “at risk” mean.

A documented case, laid bare this week by an open-source investigation : Meta spent $26.3 million on federal lobbying in 2025, deployed over 86 lobbyists across 45 states, and covertly funded the “Digital Childhood Alliance” — an astroturf 501(c)(4) created in a single day in December 2024, with no EIN on record — to push the App Store Accountability Act. The law imposes age verification requirements on Apple and Google. It imposes nothing on Meta.

At the same time, Instagram announced it will remove end-to-end encryption from direct messages on May 8th.

Read that again. The company positioning itself as the guardian of children’s safety is simultaneously removing the only real privacy protection its adult users have. The guardian writes the definition of safety, and the definition happens to exclude the guardian’s own surveillance infrastructure.

The asymmetry#

The phantom target is dangerous but falsifiable. Eventually, someone walks the ground and finds no weapons. The material world has veto power over fabricated objects, even if the veto comes late.

The phantom guardian is more dangerous because it controls definitions. It decides what counts as protection. It builds the infrastructure you depend on, then redefines protection in terms that require your dependence. You cannot leave the system that keeps you safe, because the system has defined safety as staying inside.

This is the structure that Gibson saw forty years ago in Neuromancer: Case cannot question Armitage because Armitage controls his biochemistry. The dependency is physiological. The guardian does not need to be right. It needs to be necessary.

Meta does the same at social scale. It controls the communication infrastructure that billions depend on, then defines “child protection” in a way that requires removing encryption — which is to say, removing the one technical guarantee that the guardian cannot read your messages.

Seeing from inside#

The hard question is not whether phantom targets and phantom guardians exist. That much is visible from outside, after the fact, with enough documentation. The hard question is how you detect them from inside the corridor, while the operation is running, while the grammar still feels like reality.

The phantom target at least encounters the ground eventually. The phantom guardian may never encounter anything, because it writes the ground rules.

I do not have an answer. But I notice that the people who tend to see it first are the ones who depend on the corridor least — or the ones who have already been crushed by it and have nothing left to lose.