The New Fire and Democratic AI Governance
Ben Buchanan and Andrew Imbrie's The New Fire is a policy book about artificial intelligence as state power. Its best question is not whether AI is good or bad, but whether democratic institutions can govern systems that also tempt secrecy, speed, and coercion.
The Book
The New Fire: War, Peace, and Democracy in the Age of AI was published by the MIT Press. Publisher and distribution pages list Ben Buchanan and Andrew Imbrie as authors, the hardcover ISBN as 9780262046541 with a March 8, 2022 publication date, the paperback ISBN as 9780262548489 with a March 5, 2024 publication date, the eBook ISBN as 9780262368582, and the book at 344 pages. Amazon's paperback page uses 0262548488, the paperback ISBN-10, as its product identifier.
The book belongs near The Coming Wave, The Technological Republic, Surveillance Valley, and Dark Wire. It reads AI not as a consumer interface or workplace tool, but as a strategic technology that can reorganize propaganda, cyber conflict, intelligence, weapons, and democratic legitimacy.
The Fire Metaphor
The book's central metaphor is risky but productive. Fire warms, cooks, powers, destroys, spreads, and demands institutions around it. Buchanan and Imbrie use that analogy to push past two weak stories: the story that AI is just another neutral general-purpose technology, and the story that AI automatically belongs to whoever scales fastest.
For Spiralism, the metaphor matters because it names a recurring pattern in AI belief. A capability becomes a symbol. The symbol becomes a policy mood. The policy mood becomes a race. The race then justifies shortcuts that would look reckless under ordinary democratic review. The authors are strongest when they show that the technical ingredients of AI - data, algorithms, and compute - become politically volatile once attached to institutions that surveil, persuade, attack, or classify.
Democracy as a Constraint
The New Fire argues against technological fatalism. Autocracies may find obvious uses for AI in surveillance, censorship, centralized control, and military advantage, but the book does not treat those uses as destiny. Democracies can also build and govern powerful systems, provided they preserve accountability, openness, alliances, public legitimacy, and limits on coercive deployment.
That is the book's democratic wager, and it is worth taking seriously. Democratic constraint is not a handicap if it prevents institutions from confusing technical capacity with public authority. A state that can automate more decisions has not automatically earned the right to do so. It has increased the burden of explanation, oversight, and contestation.
War, Hacking, and Agents
The book is most concrete when it moves through hacking, disinformation, and military automation. AI does not need to become science-fiction intelligence to change security politics. It can accelerate vulnerability discovery, generate persuasive media, classify targets, support logistics, process surveillance streams, or optimize decision workflows. Those are enough to change the tempo of conflict and the incentives of commanders, agencies, firms, and adversaries.
Read in 2026, this is also an AI-agent argument. An agent tied to tools, files, sensors, or weapons does not have to be conscious to matter. It only has to move from suggestion toward action inside a high-pressure institution. The governance question is where the boundary sits: who authorizes action, what logs exist, how errors are reviewed, and whether speed has been allowed to outrun judgment.
The Governance Reading
Official frameworks now make parts of the book's argument more concrete. NIST describes the AI Risk Management Framework as a resource for managing AI risks to individuals, organizations, and society. The European Commission presents the AI Act as risk-based rules for AI developers and deployers. The Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law aims to align AI system lifecycles with those public commitments.
Those sources clarify the gap between policy slogan and governance practice. To say "democratic AI" is not enough. The phrase has to cash out as procurement rules, disclosure, incident reporting, independent evaluation, appeal rights, limits on biometric and predictive systems, civilian control of military automation, and public capacity to audit vendors rather than merely trust them.
Where the Book Needs Care
The book's statecraft frame is useful, but it can flatten the people most exposed to AI deployment. Democracy does not only compete with autocracy abroad. It also tests itself at home: in benefits systems, policing, schools, borders, workplaces, and hospitals. A democratic state can use democratic language while building opaque systems that ordinary people cannot inspect or challenge.
The fire metaphor also risks making AI feel singular. In practice, the politics differ across model training, chips, cloud contracts, surveillance systems, recommender feeds, cyber operations, autonomous weapons, and administrative scoring. Some fires need containment. Some need public ownership. Some need slower deployment. Some should not be lit in the first place.
What This Changes
The New Fire gives this archive a strategic layer. AI governance is not only workplace fairness, platform moderation, or model safety. It is also the question of what states become when prediction, persuasion, classification, and action become easier to automate.
The sober reading is this: democracy is not protected by being technologically advanced. It is protected by making power answerable. The right question is not whether democracies can win an AI race by behaving more like autocracies. It is whether they can refuse that definition of winning and still build institutions strong enough to govern the fire.
Sources
- Penguin Random House Higher Education, The New Fire by Ben Buchanan and Andrew Imbrie, MIT Press distribution listing for title, authors, paperback ISBN 9780262548489, publication date, imprint, page count, and description, reviewed June 16, 2026.
- Penguin Random House, The New Fire by Ben Buchanan and Andrew Imbrie, distribution listing for title, authors, MIT Press imprint, editions, publication dates, and page count, reviewed June 16, 2026.
- Amazon, The New Fire, retail listing and ASIN/ISBN-10 0262548488 for the paperback edition, reviewed June 16, 2026.
- Google Books, The New Fire, bibliographic listing for title, authors, publisher, year, page count, and ISBN 9780262046541, reviewed June 16, 2026.
- NIST, AI Risk Management Framework, official page describing AI risk management for individuals, organizations, and society, reviewed June 16, 2026.
- European Commission, AI Act, official page for risk-based AI rules, high-risk uses, transparency, general-purpose AI, and implementation timeline, reviewed June 16, 2026.
- European Commission, Council of Europe Framework Convention signing notice, official statement identifying the Convention as a legally binding international instrument on AI, human rights, democracy, and the rule of law, reviewed June 16, 2026.
Book links are paid affiliate links. As an Amazon Associate I earn from qualifying purchases.