The Bunker Becomes the State

Naomi Klein and Astra Taylor’s End Times Fascism begins with an image so grotesque it seems designed to test the reader’s sense of reality: a gleaming, AI-rendered future city, all luxury towers, biometric access, and supervised abundance, imagined over the ruins of Gaza. The image is not offered as an aberration. It is the book’s concentrated thesis. In the political imagination of a powerful new right, devastation is no longer only a tragedy, scandal, or failure of policy. It is raw material. Ruins become real estate. Displacement becomes a security problem. The suffering of millions becomes a blank surface upon which the wealthy and powerful can project gated, technologically managed fantasies of renewal.

Klein and Taylor call this politics end times fascism. The phrase is deliberately severe, but their argument is not merely that contemporary authoritarianism resembles the fascism of the twentieth century. It is that fascism has adapted to a world its earlier forms did not know: a world of climate disruption, nuclear weapons, planetary ecological strain, mass displacement, artificial intelligence, automated surveillance, synthetic media, and wealth concentrated at a scale capable of shaping entire states.

The book’s essential insight is that a significant faction of the far right has ceased to imagine a common future worth preserving. It does not necessarily deny that the climate is destabilizing, that migration will increase, that public systems are strained, that wars can spread, or that artificial intelligence will transform social life. It often understands those realities perfectly well. Its answer is simply not collective repair.

Its answer is selection.

Who will be protected? Who will be tracked? Who will be expelled? Who will have access to water, energy, housing, health care, insurance, mobility, legal protection, and political recognition? Who will be treated as a citizen, a worker, a consumer, a security threat, a demographic threat, an administrative burden, a target, or disposable surplus?

The book’s title names an emerging politics of collapse. But its most disturbing implication is institutional. The bunker is not only a billionaire’s compound, a private city, a fortified border, a luxury development, a hardened data center, or a promised colony on Mars. The bunker is becoming a form of government.

From Disaster Capitalism to Selective Survival

Klein’s earlier work, especially The Shock Doctrine, examined how wars, financial crises, coups, terrorist attacks, and natural disasters became opportunities for privatization, deregulation, land seizure, austerity, and the transfer of public wealth upward. The central pattern was opportunism: an emergency disorients a society, then elites move quickly to impose policies that would face resistance under normal conditions.

End Times Fascism describes a darker phase.

The contemporary far right does not merely exploit shocks after they occur. It increasingly treats permanent shock as the normal condition of political life. Climate disasters, pandemics, infrastructure failures, food-price spikes, water scarcity, energy insecurity, technological upheaval, migration, geopolitical conflict, and social fragmentation are not expected to end. They are built into the imagination of the future.

This is the book’s sharpest distinction from the politics of an earlier era. Twentieth-century fascism promised a favored in-group some version of collective restoration: jobs, family order, military strength, national greatness, stolen land, social cohesion, industrial power, and material security purchased through exclusion and violence. It was monstrous, but it was still a social project. It offered parts of the population a future, however blood-soaked and fraudulent.

End times fascism offers something narrower. It offers managed insecurity.

It promises a protected zone for those deemed worth protecting: the wealthy, the armed, the technologically connected, the properly documented, the politically loyal, and the strategically useful. It offers private resilience in place of public resilience. It offers a hierarchy of shelter in place of a social contract.

Klein and Taylor describe this as a struggle over “who is going to survive the future.” Their point is not that every person associated with the political right consciously wants catastrophe. It is that powerful factions are organizing as though catastrophe is inevitable, profitable, or even spiritually redemptive—and as though the moral task is to secure the survival of a deserving remnant rather than preserve a livable world for everyone.

This is the core of selective survival.

A billionaire’s bunker, a border wall, a private fire service, a surveillance network, a biometric checkpoint, a gated smart city, a drone patrol, a data center, a luxury development built over devastation, and a government eligibility system that silently excludes people are not identical. But they can belong to the same political architecture. They divide the world between those who are insulated and those who are exposed.

The State Does Not Disappear

The anti-state rhetoric of the tech right can be misleading. It celebrates “exit,” private cities, crypto, deregulation, decentralization, innovation, disruption, freedom from bureaucracy, and escape from public constraint. It often speaks as though government itself were the obstacle to liberation.

But the governing model emerging from this politics is not anti-state. It requires an extremely capable state in selected domains.

It wants a state that can secure borders, identify migrants, build detention systems, punish disfavored groups, protect favored investment sectors, finance defense technology, subsidize energy and data infrastructure, control land, manage strategic resources, and weaken regulation. It wants rapid executive action, pliable agencies, compliant public servants, opaque contracts, and a legal environment in which private firms can become indispensable to public functions.

What it does not want is a state strong enough to restrain concentrated wealth, enforce labor rights, regulate powerful corporations, guarantee housing, sustain public health, provide durable social insurance, protect ecosystems, preserve independent expertise, or give ordinary people meaningful power over the systems that govern them.

The state does not wither away. It changes purpose.

This is where the book’s imagery of bunkers becomes politically exact. The bunker state is not necessarily a state with fewer institutions. It is a state that selectively concentrates institutional capacity in enforcement, exclusion, extraction, security, and elite insulation while allowing the institutions of common life to decay.

It may build an immense apparatus for locating people while failing to locate affordable housing.

It may develop sophisticated methods for scoring risk while failing to manage the risks created by heat, flood, disease, and poisoned water.

It may fund artificial intelligence while abandoning public systems that keep human beings alive.

It may secure data centers while permitting communities to become less secure.

The question is not whether the state is powerful. It is: powerful for whom, and against whom?

The Bureaucracy of the Bunker

Klein and Taylor are especially compelling when they describe the fantasies that animate end times fascism: rapture, space colonization, private cities, fortified nation-states, luxury refuges, and a technological transcendence from the damaged physical world. But the most consequential version of the bunker is rarely cinematic.

It is administrative.

It is a job classification that makes a career official easier to remove. It is a procurement rule that locks an agency into a private vendor’s ecosystem. It is a data-sharing agreement that joins databases once kept apart. It is a trade-secret claim that prevents the public from inspecting an automated system. It is a grant condition that pressures states to weaken regulation. It is an eligibility model that denies a benefit without providing an intelligible explanation. It is a dashboard that converts a person into a confidence score.

The administration’s creation of Schedule Policy/Career offers an example of how bureaucratic change can alter political capacity. Roughly 8,000 policy-influencing career positions were transferred into a new excepted-service category, substantially weakening incumbents’ civil-service protections and making removal easier.

That sounds technical because it is technical. It is also political.

A democratic state depends on more than elections. It depends on inspectors, scientists, attorneys, engineers, benefits administrators, public-health workers, analysts, investigators, and career professionals who can preserve expertise, administer law consistently, and resist unlawful or politically motivated pressure. If those people can be more easily removed or marginalized, the effects extend beyond any individual agency. The state becomes more vulnerable to personal command, ideological loyalty tests, and corporate influence.

This does not prove that every civil servant will be purged, that every agency has been captured, or that courts and unions no longer matter. They do. But full conquest is not necessary. A political project can advance by making independent resistance harder, public expertise scarcer, institutional memory weaker, and executive intervention easier.

The bunker is built one personnel rule at a time.

The human consequence of this architecture is not abstract. In Maryland, Kimberlee Williams spent six months in jail after police relied on a faulty facial-recognition match. An erroneous technical lead became state action, then detention, then months of lost freedom—a pattern in which responsibility can be dispersed across a system and each participant can claim merely to have relied on someone else.

Immigration as Prototype

The clearest example of technology serving this reorganized state is immigration enforcement.

Here, the abstract vocabulary of AI, analytics, data integration, operational efficiency, and workflow optimization becomes concrete. Systems identify people. Records merge. Addresses are ranked. Targets are prioritized. Officers receive leads. Detention and removal become easier to organize at scale.

Palantir’s expanding role in immigration enforcement is revealing. ICE’s longstanding sole-source relationship with the company had already exceeded $145 million before the expansion of ImmigrationOS, a platform intended to support enforcement operations, including target identification, self-deportation tracking, and removal logistics. Reporting on Palantir’s ELITE tool describes a system that pulls from government data to identify and prioritize enforcement targets, generate dossiers, and provide address-confidence assessments for people who may be subject to arrest or deportation.

These facts do not demonstrate that Palantir controls immigration policy or that software makes every enforcement decision. That would be a crude argument. The deeper concern is more ordinary and more durable.

Political decisions about whom to pursue are being translated into technical processes that outsiders may struggle to inspect, contest, or appeal.

A person can become a record. The record can become a score. The score can become a lead. The lead can become state attention. State attention can become an arrest, a detention, a removal proceeding, a lost job, a separated family, or a life reorganized by a bureaucratic error that no one institution fully owns.

The danger is not merely that computers make mistakes. The danger is that systems distribute responsibility so widely that no one remains clearly accountable when harm occurs.

In a typical data-driven enforcement chain, a government database or commercial source supplies information; a vendor may integrate it; a system may rank a lead; an officer interprets the output; and a supervisor authorizes action. An agency insists that human judgment remains involved. When the wrong person is caught in the process, each actor can point elsewhere.

The question is not only whether a system is accurate in the aggregate. It is whether a person can know they have been flagged, learn what information was used, challenge an incorrect record, obtain an explanation, and stop the state before the system’s recommendation becomes a real-world punishment.

Those are questions of democracy, due process, and human dignity. They are also questions of survival.

The Cloud Has a Body

The political world described by End Times Fascism is frequently presented as digital and futuristic: AI models, synthetic media, platforms, clouds, autonomous systems, biometric surveillance, space colonization, and virtual escape. Yet Klein and Taylor insist that the technologies of the future are radically material. Their machines require enormous quantities of electricity, water, land, minerals, hardware, cooling systems, transmission lines, and public infrastructure.

This is one of the book’s most important contributions.

AI does not float above the world. The cloud has a body.

The federal government’s rush to embed AI across agencies makes this point urgent. A tracker maintained by the Revolving Door Project reported nearly 3,000 AI uses across 29 federal agencies in 2025, an increase of more than 75 percent over 2024. Its examples included analysis of sensitive data and public-service functions, raising the prospect that tasks once carried out by public employees may be automated or supplemented. Tech Policy Press reported that the Pentagon rolled out Google’s Gemini for Government as its largest generative-AI deployment to date, while the IRS began rolling out Salesforce’s Agentforce. It also noted an executive order directing federal agencies to challenge state AI regulations deemed “onerous.”

There are defensible uses for advanced computing. Cybersecurity, medical research, accessibility, fraud detection, scientific analysis, translation, and emergency logistics can all benefit from well-governed tools. The question is not whether government should use technology. It is whether technology is transparent, auditable, privacy-protective, appealable, accountable, and proportionate to the risk it imposes.

The present direction favors speed, national-security framing, industry access, and centralized technical power. It risks producing a government reliant on systems it cannot fully inspect, explain, or replace.

That is a form of structural capture.

A contractor does not need to dictate every policy if agencies become operationally dependent on its infrastructure, personnel, data architecture, and definitions of risk. The Army’s reported ten-year enterprise agreement with Palantir, capped at up to $10 billion and consolidating about 75 existing software and data contracts into a single procurement channel, illustrates this form of dependence.

The contract is not merely a purchase. It is a decision about what the state will be able to do without a particular private company.

Fossil Tech and the Material Politics of Collapse

Klein and Taylor call the emerging convergence of artificial intelligence and extractive energy Fossil Tech. The phrase captures a reality that much AI discussion obscures: the race to build larger and more powerful computing systems requires physical expansion at precisely the moment when ecological limits are becoming harder to ignore.

Data centers require electricity. Electricity requires generation, transmission, and storage. Cooling systems require water. Hardware requires mining, factories, transport, and global supply chains. The demand for computing power can become a rationale for new gas infrastructure, expanded extraction, environmental exemptions, weakened local oversight, and public subsidies directed toward private buildout.

The data center is therefore a fitting monument to the politics Klein and Taylor describe. It is windowless, secured, energy-intensive, often distant from public scrutiny, and frequently justified in the language of national competitiveness. It promises a frictionless digital future while making new claims on the water, energy, land, and labor required for any livable present.

This is not incidental to authoritarian politics. It is part of the same hierarchy of protection.

A society that fails to secure affordable energy or water for its residents, while mobilizing extraordinary power and public support for strategically favored data centers, is making a political decision. A state that describes AI dominance as an urgent national mission while treating climate adaptation, public transit, housing, and public health as unaffordable is making a political decision. A government that subsidizes private technical infrastructure while weakening the collective systems needed to endure disasters is making a political decision.

The bunker is not self-sufficient. It is powered by the world it abandons.

The Emotional Logic of the Right

A useful aspect of End Times Fascism is its insistence that authoritarian politics cannot be explained by elite conspiracy alone. It draws strength from real fear, real precarity, real humiliation, and real institutional failure.

People facing housing costs they cannot meet, jobs that do not provide security, decaying schools, inaccessible health care, addiction, debt, crime, community disintegration, climate anxiety, and political betrayal are not necessarily seeking fascism. Most are seeking stability, dignity, order, safety, and a future.

The far right succeeds when it redirects those needs.

It says the problem is the migrant, not the landlord or private-equity firm.

It says the problem is the civil servant, not the hollowing out of public institutions.

It says the problem is diversity, not the economic order that makes social life more precarious.

It says the problem is the climate refugee, not the fossil economy accelerating displacement.

It says the problem is bureaucracy, then builds a more coercive bureaucracy behind opaque screens.

The politics of end times fascism is not sustained simply by hatred. It is sustained by an offer of relative protection. It tells people that even if they cannot be made secure, they can at least be placed above someone else. They can belong to the protected category. They can help guard the wall. They can be assured that someone else will be designated as the source of disorder.

This is why the border becomes so central. In the bunker imagination, the nation itself is a survival compound. Its walls may be made of concrete, drones, databases, militarized policing, biometric screening, detention centers, immigration courts, employment verification, and cultural panic. The purpose is not to solve the crises that force people to move. The purpose is to decide who may enter the shrinking zone of security.

The Limits of the Book

The force of End Times Fascism is also its analytical risk.

“End times fascism” gathers an enormous range of actors: religious fundamentalists, fossil-fuel interests, tech oligarchs, AI accelerationists, defense contractors, ethno-nationalists, reactionary media figures, border hawks, authoritarian politicians, and private-equity opportunists. Their interests overlap, but they do not form a seamless bloc.

They disagree about religion, the role of the nation-state, the value of global markets, foreign war, artificial intelligence, migration, labor, and even the boundaries of the category they want to protect. The conflict between Christian nationalism and Silicon Valley libertarianism is real. The conflict between isolationist rhetoric and militarized foreign policy is real. The dependence of technology companies on skilled immigration, global supply chains, stable energy systems, research universities, and consumer confidence creates tensions with the nationalist politics they may support.

The book knows this. But its language sometimes moves so quickly from affinity to unity that readers may lose sight of the fractures.

Those fractures matter. They are openings for resistance. They make the coalition unstable. They can produce policy failures, public disillusionment, elite conflict, and political defections.

But instability does not make the project harmless.

Fascist coalitions have often been incoherent. Their power lies not in perfect coordination but in their ability to converge around common enemies, common fears, and common permissions. They do not need to agree on the final destination if they agree that democracy is an obstacle, that equality is weakness, that the vulnerable are expendable, that public regulation is intolerable, and that the future belongs to those willing to seize it.

The shared principle is impunity.

A Realistic Picture of Collapse

The most realistic danger is not a seamless, all-powerful technocratic regime. It is a brittle and unequal order.

It may be technologically advanced in protected sectors. It may possess sophisticated tools for surveillance, targeting, classification, and enforcement. It may direct immense resources toward defense, data infrastructure, private security, and extraction. It may become more efficient at deciding who is eligible, who is risky, who can be expelled, and who can be ignored.

Yet it may become progressively less capable of maintaining the ordinary conditions of collective life.

It may build data centers while housing becomes unaffordable.

It may militarize borders while public-health systems deteriorate.

It may develop tools for predictive enforcement while water systems fail.

It may centralize executive power while local capacity to respond to heat, fire, flood, and disease erodes.

It may offer the wealthy air conditioning, private medicine, insurance, security, and mobility while offering everyone else paperwork, waiting lists, policing, debt, and exposure.

This is what collapse looks like in a rich society. It does not arrive as a single event. It is distributed unevenly.

It appears first as an insurance policy canceled after a flood. A hospital closure. A heat wave without a cooling center. A school without staff. A public agency that cannot answer the phone. A water bill no household can afford. A benefits system that denies a person through an automated notice. A city that can support a corporate data campus while struggling to repair its pipes. A community that is told there is no money for public housing but discovers abundant money for detention, surveillance, and security contracts.

The wealthy can purchase distance from these conditions. They cannot purchase escape from the social and ecological systems upon which their wealth depends.

Private security cannot produce a stable public health system.

A bunker cannot repair a grid.

An algorithm cannot restore trust.

A gated city cannot replace the people who build, maintain, transport, teach, heal, grow food, clean, care, and govern.

The fantasy of exit is, in that sense, not only cruel. It is delusional.

The Counter-Architecture

Klein and Taylor’s answer is “hereness”: an ethic and politics of staying with the living world rather than escaping from it. They oppose the bunker fantasy with rootedness, mutual obligation, solidarity, and the practical work of sustaining communities through crisis.

That answer should not be reduced to sentiment.

Hereness requires a counter-architecture.

It requires public institutions capable of providing housing, health care, clean water, food security, affordable energy, climate adaptation, public transit, emergency response, education, labor protections, and due process. It requires a civil service insulated from partisan purges and corporate pressure. It requires procurement systems that prevent government agencies from becoming permanently dependent on opaque private vendors. It requires technology that can be inspected, challenged, explained, and governed in the public interest.

It requires a politics that treats resilience as universal rather than exclusive.

The opposite of the bunker is not naïve faith in an old liberal order that has already failed many people. It is the construction of institutions capable of making survival less conditional on wealth, status, race, citizenship, location, and political usefulness.

The future will contain shocks. Climate disruption is already here. Migration will continue. Technology will change work, culture, and governance. Institutions will fail unless they are rebuilt. No serious politics can deny these facts.

The question is what kind of society will be built around them.

Will crisis become an excuse to classify people into protected and disposable groups?

Will technology be used to deepen private power and public dependency?

Will the state become a bunker—efficient at enforcement and incapable of care?

Or will the recognition of shared vulnerability force a more democratic, materially grounded, and humane politics?

End Times Fascism is valuable because it insists that these are not abstract questions. They are being answered now, in borders, contracts, personnel systems, data centers, weapons platforms, immigration databases, energy policy, and the slow weakening of the public institutions that make collective survival possible.

The architecture of control is already under construction.

The task is to prevent it from becoming the normal shape of the future.