The Doctrine of Expropriation: A Political Blueprint for Reclaiming the Commons
When we speak of “expropriation,” we must discard the crude image of a midnight raid on corporate headquarters. Instead, think of it in its broadest sense: the decisive transfer of ultimate control and benefit from a private oligarchy to the public domain. This is the political commitment required to treat foundational artificial intelligence capabilities not as a speculative commodity, but as a utility—a digital grid or an interstate highway system that is too essential to be left solely to the caprice of profit motives.
Moving Beyond Regulation to Operational Control
Incremental regulation—the familiar approach to manage immediate harms like bias, misinformation, or market manipulation—is fundamentally insufficient. Why? Because the concentration of power *itself* constitutes the primary, existential threat. Regulations are rules *for* the market; operational control is the act of *replacing* the market’s prerogative in a domain critical to sovereignty. The political focus must pivot to establishing mechanisms where the state actively manages the direction, deployment, and intellectual property inheritance of frontier AI. This necessitates demanding a clear, non-negotiable path for the state to assume ownership or mandatory operational custodianship of models deemed critical to national security the moment they are completed, or the instant a defined financial distress event is triggered. The oligarchy, in their haste to scale and secure defense dollars, is effectively selling the tools of their own containment. The state’s duty is simply to recognize this transaction and act upon it before the inevitable market correction arrives. This is about moving from a posture of *asking* for cooperation to one of *enforcing* sovereignty over strategic technological assets.
The Case Against Private Custodianship of Strategic Assets
The core, unassailable argument against the current arrangement is that strategic assets—those upon which the nation’s defense, security, and fundamental economic functionality depend—cannot be solely entrusted to entities whose fiduciary duty is defined by quarterly earnings reports. The comparison to nuclear technology is unavoidable, but focus on stewardship, not just creation. The nation would never entrust the entirety of its nuclear deterrent command and control systems to a handful of private investors whose stock options were contingent on rapid deployment, irrespective of long-term safety or strategic stability. Artificial intelligence, with its reasoning capabilities and its increasing control over automated systems, is rapidly becoming that exact kind of asset. To allow private interests to dictate:
- The pace of development (driven by investment cycles).. Find out more about Applying Defense Production Act to control AI infrastructure.
- The cost of access (driven by maximizing revenue).
- The ethical guardrails (often ignored in the race for the next performance benchmark).
- Advanced climate modeling and mitigation strategies.
- Optimizing public health infrastructure and resource allocation.
- Creating truly impartial, accessible educational tools for every citizen.. Find out more about Applying Defense Production Act to control AI infrastructure overview.
- Factor in the Hidden Costs: Recognize that the true cost of frontier AI is not just the training budget but the strain on energy grids and water resources—a structural weakness that requires public oversight.
- Understand the Legal Leverage: The Defense Production Act remains the most direct legal mechanism for asserting priority control over critical component supply chains and production mandates, a tool already being considered in the current climate.
- Track the Capital Concentration: Monitor the actions of the Pentagon’s Office of Strategic Capital, as its growing role in directing private capital is the clearest sign of the government’s dependency on, and therefore potential leverage over, the private AI sector.
- Demand a New Mandate: The political fight is not over regulating *what* AI does, but controlling *who* decides *what* AI *can* do. Focus on establishing mechanisms for the state to assume operational custody of strategically critical models upon threat of insolvency or completion.
…is an abdication of the state’s primary responsibility: the protection and preservation of the public sphere. The risk of societal dependence—where the nation’s operational security becomes beholden to the shifting business interests of a few AI executives—is too high to tolerate any arrangement short of ultimate public accountability. This conversation about public accountability is essential for anyone tracking the governance of frontier AI.
Historical Parallels and Political Precedents for State Intervention. Find out more about Applying Defense Production Act to control AI infrastructure guide.
To dismiss the necessity of intervention on this scale as “radical” is to suffer from a profound historical amnesia. The creation of the modern American economy—its industrial capacity, its transportation backbone, its technological dominance—was predicated on powerful, decisive acts of the state reshaping private enterprise for the common good, especially during moments of technological transition or existential challenge.
Lessons from Wartime Industrial Mobilization and Infrastructure Seizures
Throughout the twentieth century, the federal government asserted immense, necessary power over industrial capacity. From the mobilization for World War I and World War II, which saw the federal government dictating production schedules, to the rapid infrastructure deployment during the Cold War, these interventions were seen as feats of national organization, not ideological overreach. The Korean War, which birthed the DPA, offers the most direct historical blueprint for controlling the production of critical components. Furthermore, in earlier eras, the state assumed control over essential monopolies like telegraphs and railroads when their utility to the public definitively superseded their private profit potential. The twenty-first century challenge is perfectly analogous: the complex, opaque network of advanced computation and data infrastructure has become the new essential utility. Historical precedent dictates that when private interests fail to guarantee public access and security, the state *must* step in to manage the infrastructure itself. This involves drawing on the operational playbook refined during those earlier periods of state-led economic direction.
Contrasting State Power with Venture Capital’s Short-Term Horizon
The core friction between the state and the AI oligarchy is fundamentally a clash of time horizons. Venture capital, by its very nature, operates on a five-to-seven-year cycle, optimizing for rapid scaling and a lucrative exit event. The state, conversely, must operate on a fifty-year or multi-generational horizon, concerned with enduring stability, long-term technological superiority, and the systemic resilience of our democratic institutions. When the technology in question is a platform upon which all future national power—economic, military, and social—will likely rest, the short-term financial calculus of the venture capitalists becomes an active liability to the nation’s long-term survival. By assuming operational control through mechanisms like the DPA, the state can enforce a longer view, prioritizing foundational safety, verifiable alignment, and equitable access over the immediate maximization of shareholder value. This is not an act designed to crush innovation; it is an act designed to reorient the *direction* of innovation toward durable public goods rather than toward the creation of speculative private monopolies.
The Infrastructural Nexus: Data, Energy, and the Control of the Physical World. Find out more about Applying Defense Production Act to control AI infrastructure tips.
The algorithms and the models themselves are merely the visible, glittering tip of the technological iceberg. True, enduring control over the AI complex resides in the mastery of the physical resources required to train and run these colossal models: the data centers, the fabrication plants for the chips, the immense power grids feeding them, and, critically, the water needed for cooling. The oligarchy’s power is fundamentally rooted in their control over this often-overlooked physical substrate.
The Hidden Costs: Water Consumption and Data Center Footprints
The enormous, geographically concentrated computational power demanded by leading AI labs places an unsustainable strain on local resources, frequently in regions already stressed by climate volatility. The cooling systems for these massive server farms require prodigious amounts of water, placing them in direct competition with agricultural and municipal needs—a fact often obscured by non-disclosure agreements related to defense contracts or sheer corporate secrecy. A true sovereignty strategy must map these physical dependencies. Control over the physical loci of computation—the land deeds, the long-term energy contracts, and the water rights supporting these data centers—is perhaps the most tangible and immediate point of leverage the state possesses. As evidence shows that energy demand from AI is creating structural shifts in the power market, the state must use its regulatory authority over public utilities and environmental stewardship to impose conditions that effectively transfer operational control over these vital choke points, regardless of the legal ownership of the intangible software running above them.
Mapping the Supply Chain Dependencies for State Acquisition
The path to exerting state power requires a meticulous audit of the entire AI supply chain, a task the government is already beginning to formalize. This chain stretches from the fabrication plants producing specialized chips—which the White House’s July 2025 Action Plan identifies as a priority for securing frontier AI development—to the logistics networks delivering cooling infrastructure. The oligarchy is deeply integrated into global supply chains, but these chains are subject to national jurisdiction at key logistical nodes. If the state cannot nationalize the software, it can certainly commandeer the *means of its reproduction and maintenance*. Directives under the DPA could immediately prioritize the production of necessary semiconductors for state-controlled AI development, impose radical transparency requirements on all energy providers serving these facilities, and mandate real-time access for government inspectors to audit energy and water consumption. By asserting control over the physical means of sustaining the AI enterprise, the state can effectively starve or redirect the private sector’s operations, making state direction the *de facto* path of least resistance for the entire complex.
Envisioning the Post-Oligarchic AI Future: A Democratic Dividend
The purpose of this intervention is not merely punitive or defensive; it is fundamentally constructive. Reclaiming control from the hands of a few speculators allows the nation to redirect this immense technological power toward universal societal benefit, transforming a speculative commercial engine into a true public servant.
The Creation of Public-Sector AI Models for National Benefit. Find out more about Applying Defense Production Act to control AI infrastructure strategies.
With the foundational models and computational infrastructure secured under clear state direction, the immediate priority must be the rapid development and deployment of openly accessible, public-sector foundational models. These models, trained and governed by standards designed for transparency, reliability, and the public good—rather than clandestine optimization for engagement metrics or defense contracting specifications—would represent a true democratic dividend. Imagine the vast resources currently poured into proprietary, closed systems being redirected toward models focused specifically on:
This public alternative immediately breaks the oligarchy’s monopoly on advanced reasoning, allowing genuine, decentralized innovation to flourish in a competitive, yet democratically accountable, ecosystem. The intellectual property generated through state subsidy and control would be treated as a national resource, released under liberal licensing to foster widespread, diverse application across civil society—from small businesses to local government services.
Reorienting Research Away from Market Maximization to Societal Resilience
Finally, state intervention serves to recalibrate the entire enterprise of artificial intelligence research. As long as the dominant financial incentive structure favors systems that maximize proprietary advantage or cater only to the most powerful military or commercial clients, the most pressing societal needs—those that lack an immediate, massive revenue stream—will be systematically ignored. The state’s control allows for the imposition of a new research mandate, one centered on resilience, verifiable alignment, interpretability, and long-term societal integration, rather than simply achieving the next benchmark in performance metrics that only the oligarchy’s peers can even verify. This reorientation acknowledges a fundamental truth: the true measure of artificial intelligence success in the twenty-first century will not be its speed or its complexity, but its trustworthiness, its alignment with enduring human values, and its demonstrable contribution to a more stable and equitable world. This vision is entirely incompatible with the profit-driven imperatives of the current algorithmic overlords. The path forward is not one of gentle persuasion; it is a deliberate act of political will to reclaim the future being sold off in pieces by the few. ***
Key Takeaways and Actionable Insights (As of November 27, 2025)
The economic volatility of the AI sector, driven by exponential cost growth and massive resource consumption, creates an unacceptable national security risk. Here is what the current structure demands of clear-eyed observers:. Find out more about Legal pathways for state expropriation of frontier AI definition guide.
The time for merely observing the speculative sprint is over. The state must act as the ultimate steward of this foundational technology. What actions do you believe are necessary to enforce a multi-generational time horizon on this critical infrastructure, moving innovation away from pure market maximization?
For further reading on the economic realities underpinning this argument, see the analysis on the rising costs of training frontier AI models and the implications of the White House AI Action Plan released in July 2025.
