The Tech Coup and the Outsourcing of Democratic Power
Marietje Schaake's The Tech Coup: How to Save Democracy from Silicon Valley is a book about democratic institutions losing operational power to private technology companies. Its subject is not only Big Tech influence. It is the transfer of public functions into systems that citizens cannot vote out, inspect on ordinary terms, or govern through familiar constitutional channels.
The Book
The Tech Coup was published by Princeton University Press in 2024. McKinsey's author interview lists the publication date as September 24, 2024, the hardcover at 336 pages, and ISBN 9780691241173. JSTOR records the same book under Princeton University Press with the DOI 10.2307/jj.13359161 and a table of contents moving from "The Code" and "Tech on the Front Lines" toward "Reclaiming Sovereignty" and "Prioritizing the Public."
Schaake writes from unusual institutional proximity. Stanford HAI identifies her as a Stanford Cyber Policy Center and Institute for Human-Centered AI fellow, a Financial Times columnist, a former Member of the European Parliament from 2009 to 2019, and a member of the UN High-Level Advisory Body on AI. That background shapes the book: it is less a platform memoir than an argument from someone who has watched democratic institutions try to regulate systems they also increasingly depend on.
The book belongs beside The Whale and the Reactor, The Master Switch, The Black Box Society, Surveillance Valley, and The Stack. Its central question is practical: what happens when core democratic capacities move into private technical systems faster than law, public administration, and civic understanding can follow?
What Makes It a Coup
The title is deliberately sharp. Schaake is not claiming that Silicon Valley tanks have rolled into parliament. She is naming a slower transfer of authority: companies build systems that become infrastructural, governments adopt them or struggle to regulate them, and then public institutions find that their own capacity has been displaced.
That is why the strongest word in the book is not "innovation" or "disruption." It is outsourcing. Public agencies outsource cloud infrastructure, identity systems, cyber tools, communications platforms, data analytics, content moderation pressure, battlefield connectivity, procurement expertise, and policy imagination. Each contract may look narrow. Together they change who can act, who understands the system, and who can say no.
This is a familiar pattern in digital life. A service becomes convenient. Then it becomes standard. Then it becomes the practical condition for participation. By the time a public argument catches up, the question is no longer whether the system should govern the domain. The question is how to bargain with a system everyone already relies on.
State Functions Without State Constraints
Stanford HAI's discussion of the book summarizes Schaake's examples as companies moving into areas such as cybersecurity, policing, elections, and military defense policy. The concrete cases matter because they show that "tech policy" is no longer a narrow consumer-protection field. It is about the operational machinery of sovereignty.
Spyware is the clearest case. A private company can sell intrusion capabilities that resemble intelligence-service powers, while the people targeted may include journalists, dissidents, judges, political opponents, or civil-society actors. Offensive cyber firms raise a related problem: private actors can cross borders in the name of defense while democratic oversight lags behind the actual capability.
Platforms also shape public speech and political legitimacy. They decide what is amplified, removed, labeled, monetized, demonetized, archived, or forgotten. They may cooperate with states, resist states, or substitute their own policy systems for public law. The public experiences this through interfaces; governments experience it as dependency; companies experience it as product governance.
The most important point is that private technical power often lacks the ordinary disciplines of public authority: due process, transparency, procurement accountability, democratic contestation, archival duties, equality obligations, and the expectation that coercive power must justify itself to the people it governs.
The AI-Age Reading
Read in 2026, The Tech Coup is an AI governance book even when it is not only about AI. Foundation models, agent platforms, cloud compute, model APIs, safety evaluations, content provenance tools, education software, workplace copilots, and public-service chatbots all intensify the same dependency problem.
AI systems are not merely products once they enter courts, classrooms, clinics, agencies, police departments, political campaigns, welfare offices, and military planning. They become institutional cognition. They summarize evidence, route cases, draft policy, classify risk, generate public messages, monitor behavior, and shape what decision-makers believe they know.
This creates a recursive governance problem. Public institutions need technical capacity to govern AI. But if that capacity is supplied by the same private systems being governed, the state rents its own ability to see. The dashboard, model card, audit interface, cloud log, procurement memo, and vendor explanation become the official picture of reality. Oversight can then become a supervised tour through someone else's machine.
Schaake's book is strongest when read as a warning against that loss of independent capacity. Democratic control cannot mean asking powerful vendors to behave responsibly while public institutions forget how the systems work. It has to include public expertise, adversarial testing, procurement leverage, enforceable rights, contestability, exit plans, and institutions able to operate outside the vendor's preferred story.
Where the Book Needs Pressure
The book's framing is useful, but it can sound more unified than the problem really is. "Silicon Valley" is a convenient label, while the power landscape includes cloud providers, chip firms, AI labs, defense contractors, spyware vendors, telecom networks, app stores, payment systems, data brokers, consultancies, open-source ecosystems, and state-aligned technology companies outside the United States.
There is also a risk in treating the state as the automatic remedy. States can be captured, authoritarian, underfunded, technically weak, or eager customers for surveillance and control. A democratic answer cannot simply move power from companies to governments. It has to rebuild public constraints around both: courts, journalism, civil society, labor power, libraries, auditors, standards bodies, international law, and meaningful rights for the people processed by these systems.
Finally, the book's policy road points need operational follow-through. "Democratic control" becomes real only when a procurement officer can reject a vendor lock-in clause, a teacher can refuse a classroom system without punishment, a benefits recipient can appeal an automated decision, a researcher can inspect a model's public claims, and a city can leave a platform without losing its institutional memory.
The Site Reading
The practical lesson is that technology governance must be evaluated as institutional capacity, not just rule writing.
Ask who can understand the system without vendor permission. Ask who can audit it without marketing mediation. Ask who owns the logs, categories, models, data flows, and institutional memory. Ask whether citizens have appeal rights that reach a human authority with power to override the machine. Ask what happens if the provider changes terms, fails, is acquired, becomes politically hostile, or is asked to serve two governments with conflicting demands.
This is where The Tech Coup connects to recursive reality. A private system does not only execute policy. It supplies the categories through which institutions perceive the world. Those categories guide action. The resulting action produces new data. The data confirms the system's usefulness. Over time, the technical map becomes administrative common sense.
Schaake's enduring warning is that democracy can lose power without a single dramatic surrender. It can lose power contract by contract, dashboard by dashboard, integration by integration, until public authority still has the formal title but no longer owns the machinery of action.
Sources
- JSTOR, The Tech Coup: How to Save Democracy from Silicon Valley, Princeton University Press bibliographic record, DOI, copyright year, and table of contents, reviewed May 19, 2026.
- McKinsey Global Publishing, "Author Talks: Making tech work for us", author interview with publication date, publisher, page count, and ISBN, November 14, 2024, reviewed May 19, 2026.
- Stanford HAI, "The Tech Coup: A New Book Shows How the Unchecked Power of Companies Is Destabilizing Governance", October 7, 2024, reviewed May 19, 2026.
- Stanford HAI, Marietje Schaake profile, author roles and policy background, reviewed May 19, 2026.
- Liz Scheier, Stanford Social Innovation Review, "The Great Outsourcing of Government", review of The Tech Coup, Fall 2024, DOI 10.48558/FEQ8-D283, reviewed May 19, 2026.
- Kirkus Reviews, The Tech Coup, review posted May 23, 2024, issue date July 1, 2024, reviewed May 19, 2026.
Book links are paid affiliate links. As an Amazon Associate I earn from qualifying purchases.
- Amazon, The Tech Coup by Marietje Schaake.