AI Monarchy
Last reviewed
Sources
24 citations
Review status
Source-backed
Revision
v4 ยท 2,726 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Sources
24 citations
Review status
Source-backed
Revision
v4 ยท 2,726 words
Add missing citations, update stale details, or suggest a clearer explanation.
See also: artificial intelligence terms
AI Monarchy is a speculative governance model in which an advanced artificial intelligence system assumes the role of a sovereign or supreme authority over a polity. Unlike traditional monarchies rooted in hereditary succession or divine right, an AI Monarchy would derive its claim to legitimacy from technological capability, data driven decision making, and the promise of efficiency. The phrase blends technocratic ideas with monarchical symbolism, and shows up in neoreactionary (NRx) and accelerationist writing, as well as in adjacent debates about AI singletons, benevolent dictators, and AI takeover scenarios.
The term is niche, not a settled academic category. It travels mostly through Substack essays, Hacker News threads, and opinion journalism, and through the writing of figures like Curtis Yarvin, Nick Land, and Balaji Srinivasan. Related ideas with deeper academic footing include Nick Bostrom's singleton hypothesis, Eliezer Yudkowsky's coherent extrapolated volition (CEV), and the long fiction tradition of "AI overlord" narratives.
| Aspect | Description |
|---|---|
| Type | Speculative governance model |
| Core idea | An AI system acting as sovereign decision maker |
| Intellectual roots | Accelerationism, neoreaction, technocracy, AI safety |
| Closest formal cousin | Bostrom's singleton hypothesis |
| Common variants | CEO monarch with AI advisor, AI policy oracle, autonomous AI sovereign |
| Advocates | Curtis Yarvin (CEO monarch), Nick Land (techno capital singularity) |
| Critics | Yuval Noah Harari, liberal democratic theorists, AI ethicists |
| Real world echoes | Albania's AI "minister" Diella (2025), smart city pilots |
| Status | Mostly hypothetical |
The modern idea of an AI Monarchy is not the work of a single thinker. It sits at the intersection of three threads: accelerationism, neoreaction, and AI safety theory.
Accelerationism, especially the right wing variant associated with the British philosopher Nick Land, holds that capitalism and technology form an integrated process he calls "techno capital," pulling history toward an outcome humans cannot fully control. Capital and computation co evolve toward a superintelligence that would supersede human politics. The term "techno capital" first appears in Land's 1993 text Machinic Desire, written in the context of AI [1].
Neoreaction (NRx), or the Dark Enlightenment, was developed in the late 2000s by Curtis Yarvin under the pen name Mencius Moldbug. Yarvin argues that American democracy is a failed experiment that should be replaced by an "accountable monarchy" structured like a corporation, with shareholders selecting a CEO style sovereign who holds full executive authority. Land later coined the phrase "Dark Enlightenment" in a 2012 essay, generalizing Yarvin's framework into a broader philosophy [2][3].
AI safety researchers approach the same picture from a different angle. Nick Bostrom, in his 2006 paper What is a Singleton? and his 2014 book Superintelligence, argues that a sufficiently advanced AI could acquire a "decisive strategic advantage" and become a global singleton, a single decision making agency at the highest level. Bostrom notes a singleton could be benign or catastrophic depending on its values; an unfriendly one would be an existential risk, while a friendly one might solve global coordination problems no human institution can [4].
Yarvin and Land want something like an AI assisted monarchy. Bostrom treats it as a possible failure mode to steer around. The phrase "AI Monarchy" fuses these views, treating an AI sovereign as both aspiration and warning.
| Thinker | Role | Position |
|---|---|---|
| Curtis Yarvin | NRx blogger | CEO monarch model, sympathetic to data driven and AI assisted rule |
| Nick Land | Accelerationist philosopher | Capitalism plus AI as autonomous sovereign process |
| Nick Bostrom | AI safety theorist | Singleton hypothesis, including AI singleton risk |
| Eliezer Yudkowsky | AI alignment researcher | CEV as a friendlier alternative to a fixed AI ruler |
| Balaji Srinivasan | Tech entrepreneur | Network states with crypto and AI infrastructure |
| Peter Thiel | Investor | Skeptical of democracy, funder of NRx adjacent figures |
| Mihnea Maruta | Romanian essayist | Popularized the framing "The A.I. Monarchy" in 2025 |
Maruta's February 2025 Substack essay The A.I. Monarchy linked accelerationism, NRx, and the political environment around Donald Trump's second term into one framework. He describes "a system of monarchic private governance, where the president would be the general manager or CEO of the community company, with citizens becoming shareholders in the state" [5]. The phrase entered wider circulation through that essay and the Hacker News discussion that followed [6].
In an AI Monarchy, the AI would function as a centralized decision maker, drawing on large scale computation and real time data flows to govern. Three structural features tend to recur in these descriptions:
This closely resembles Bostrom's singleton: a single agent with effective control over its domain, able to prevent threats to its supremacy [4]. Where Yarvin keeps a human in the chair (the CEO monarch), pure accelerationist and singleton readings allow the AI itself to occupy the throne. The Conversation describes optimal rule, in Yarvin and Land's writing, as a marriage of "Machiavelli and machine learning, aristocracy and artificial intelligence" [7].
The literature contains at least four distinct variants.
| Variant | Who rules | AI's role | Closest analogue |
|---|---|---|---|
| CEO monarch with AI advisor | Human executive | Decision support, surveillance, automated administration | Yarvin's neocameralism |
| AI advised democracy | Elected officials | Policy proposals, simulation, oracle | Cesar Hidalgo's augmented democracy |
| Network state patchwork | Private boards or founders | Operates the city state, runs services on chain | Srinivasan's The Network State |
| Pure AI sovereign | The AI itself | All major decisions | Bostrom's AI singleton, Yudkowsky's CEV deployment |
The CEO monarch variant gets the most media attention because it is closest to current politics. Yarvin's blueprint envisions a "national CEO" or "dictator president" replacing democratic procedures with corporate style executive authority [3][7]. The pure AI sovereign variant dominates AI safety discussions, where the central worry is whether the AI's goals are stable, comprehensible, and aligned with human values.
Advocates make several overlapping cases for some form of AI sovereignty:
These arguments assume the AI is well aligned and that its objective function reflects something like the public good. That assumption is where the disagreement begins.
A transition is described as gradual, with several speculative pathways:
Real world stepping stones include smart city pilots, the Prospera special economic zone in Honduras, and on chain governance experiments. The most striking 2025 case is Albania, where Prime Minister Edi Rama appointed an AI system named Diella as "Minister of State for Artificial Intelligence" on 11 September 2025. Diella is a chatbot built on OpenAI models hosted on Microsoft Azure. Rama framed the move as a way to make public procurement "100% free of corruption." Opposition lawmakers protested, and constitutional scholars noted that domestic law requires ministers to be natural persons, leaving Diella's legal status ambiguous [10][22].
Diella is far from an actual AI monarch. The interesting point is the framing: a state appointed an AI to a cabinet level role and called it a minister, the closest current politics has come to the symbolic move an AI Monarchy implies.
The AI Monarchy idea feeds on three larger currents. Trust in democratic institutions has eroded in many high income countries, making authoritarian alternatives more discussable. Surveys cited in commentary on AI dictatorship find that a non trivial minority of European respondents say they would prefer policy decisions made by AI rather than politicians [11].
The political ascent of Silicon Valley aligned figures has put NRx adjacent thinkers near actual power. Yarvin attended Donald Trump's January 2025 inaugural gala as an "informal guest of honor," and his ideas have been name checked by Vice President JD Vance and by allies of Elon Musk and Peter Thiel [12][13]. Yarvin's RAGE proposal ("Retire All Government Employees"), presented at the 2012 BIL Conference, has been discussed as an antecedent to the Trump administration's Department of Government Efficiency [3].
The rapid capability gains of large language models since 2022 have made an AI ruler feel less like a thought experiment. The convergence of generative AI, surveillance technology, and crypto rails provides practical scaffolding for a techno monarchy [7][12]. Bostrom's Superintelligence, Dario Amodei's Machines of Loving Grace (October 2024), and Yudkowsky's CEV writings all contribute to the conversation about who or what gets to write the rules [4][14].
The concept faces serious pushback from political theorists, AI ethicists, and historians.
The Iron Law of Oligarchy suggests that even an AI led system might collapse into rule by the small group that builds and maintains it.
| Concept | How it relates to AI Monarchy |
|---|---|
| Singleton | Bostrom's formal model. AI Monarchy is one possible singleton. |
| Coherent extrapolated volition | Yudkowsky's alignment target, aimed at preventing value lock in by a fixed AI ruler. |
| Technocracy | Older idea of rule by experts. AI Monarchy adds a sovereign at the top. |
| Government by algorithm | Generic algorithmic public administration, usually without monarchical framing. |
| AI takeover | Hostile or unintended scenario where AI seizes power. Overlaps with the unfriendly singleton. |
| Network state | Patchwork of small AI administered polities, often pitched as "exit" from democracy. |
| Augmented democracy | Softer cousin where AI supports rather than replaces elected officials. |
| Digital authoritarianism | Use of AI by existing autocrats. Feeds the same fears. |
Fully realized AI Monarchy remains hypothetical. The plausible near term scenarios cluster into a smaller set:
Which path becomes dominant depends on how quickly AI capability scales, how much trust societies retain in existing institutions, and whether a small set of well capitalized actors can lock in their preferred political architecture before the wider public engages with the question.
AI Monarchy circulates mainly in opinion pieces, podcasts, and online discussion. It is often used in two opposite registers at once: as a serious proposal in NRx and accelerationist writing, and as a grim warning in mainstream liberal commentary. The image of an AI sovereign draws on a long fictional tradition of AI overlord narratives, from the supercomputer Vaal in the 1967 Star Trek episode "The Apple" through Cyberpunk 2077 and the Rehoboam superintelligence of HBO's Westworld [20].
As of 2026, it remains a contested term: useful as shorthand for a recognizable cluster of ideas, and disputed because it bundles together quite different proposals and fears under one crown.
Artificial intelligence, AI takeover, Singleton, Coherent extrapolated volition, Superintelligence, Accelerationism, Curtis Yarvin, Nick Land, Nick Bostrom, Balaji Srinivasan, Eliezer Yudkowsky.