Government by machine, instead of by consent.
Jill Lepore is a Harvard historian and a New Yorker staff writer, and this is her attempt to give a name to whatever has been quietly replacing liberal democracy. The Artificial State, in her account, is a successor form: public conversation owned by private companies, public functions handed to machinery, and legitimacy claimed from data rather than from votes.
It presented itself as futurism. Jill Lepore
to liberal democracy.
A machine cannot consent on your behalf. That is the whole objection, and everything else in the book is its history.
What she means by artificial.
The Artificial State is not a metaphor for surveillance and it is not simply a complaint about social media. Lepore uses it to describe a form of rule with specific properties: the public square is privately owned, so the terms of political speech are set by companies rather than constitutions; the work of governing — allocating, scoring, deciding, predicting — is increasingly performed by systems; and the justification for a decision becomes the output of a model rather than an argument anyone has to win. Consent, the thing a liberal state is supposed to run on, becomes decorative.
This did not start with the smartphone.
The move that makes this a history book rather than a tech polemic is the timeline. Lepore reaches back centuries — to the era when political theory first described the state itself as an artificial body, an engineered person assembled out of real ones, and when the corporation was invented as another artificial person with rights but no mortality. Once you have been taught to accept that a fiction can hold power, delegating judgement to a mechanism is a much shorter step than it looks. The argument is not that the seventeenth century caused the present. It is that the present is the latest chapter of a habit, and habits are harder to break than inventions.
Arendt wrote the template. Lepore updates it.
The book is openly modelled on Hannah Arendt’s The Origins of Totalitarianism, and knowing that tells you how to read it. Arendt’s method was to show that a political catastrophe had long, ordinary, traceable roots — imperialism, bureaucracy, statelessness, mass loneliness — rather than arriving as a sudden madness. Lepore applies the same method to a softer transformation, one with no coup and no single villain, where the change is legible only in aggregate: which institutions stopped being able to decide things, and which systems quietly started.
Nobody announced the change of regime. The forms stayed exactly where they were; only the deciding moved.
The Mdrn UrbanFrom persuading voters to modelling them.
Lepore has been working this seam for years — her earlier book If Then recovered the 1960s company that first tried to simulate the American electorate — and here the thread reaches its conclusion. A campaign that can model an electorate stops needing to convince it: the task becomes finding the segments that move, addressing each with a different message, and leaving the rest unaddressed. What disappears in that transition is the thing elections were supposed to produce, which is a public argument that everyone has heard. A hundred million private messages are not a debate, and a party optimised for prediction has no particular reason to produce one.
Built by people with no theory of government.
One of the sharpest passages concerns where the architects of this order got their ideas. Not from political philosophy, constitutional history or the study of institutions, but from novels and films — the futures they read as teenagers, imported wholesale as design briefs. Hence the recurring pattern Lepore identifies: what is presented as futurism keeps resolving into something very old, because an engineer improvising a theory of legitimacy under deadline tends to reinvent, without noticing, the arrangements that political thought spent centuries learning to refuse.
The title contains a second half.
The last movement is the reason the book is not despair. Lepore’s position as a historian is that history is made by people, societies and ideas — not by machines, and not by any technology’s internal logic. Whatever was assembled through argument, capital and legislation can be disassembled the same way; the sense of inevitability is itself a claim, and one of the more useful things the Artificial State produces on its own behalf. Whether you find the ending persuasive or a little willed, the demand is concrete enough to act on: stop treating the arrangement as weather.
Anything built by argument can be argued down.
That is the sentence to carry out of this book. The most effective thing any system of power does is to make itself look like a fact about the world rather than a decision somebody took — and the historian’s job, which is also the citizen’s, is to keep producing the receipts that show it was a decision.
Summarised in my own words from The Rise and Fall of the Artificial State by Jill Lepore (Allen Lane, UK / Liveright, US; published August 2026, 336 pages). Longlisted for the Financial Times business book award. Her earlier If Then covers the prehistory of political data modelling.