Restricted Data: The History of Nuclear Secrecy in the United States

by Noah Zachary
published on 28 July 2025

Both readable and rigorous, this book offers an invaluable overview of how the classification scheme around nuclear technology evolved -- and fruitfully critiques its effectiveness.

Democracy and security have never gotten along in American society; even as nearly everyone agrees that some information should be kept from the public, just as many agree that secrecy opens the door to corruption, tyranny, and conspiracy. Perhaps there is no better example of this tension than the nuclear establishment, where the stakes of failure are measured in millions of deaths. But despite the risks, nuclear secrecy is subject to constant suspicion – and, for the most part, is riddled with breaches and leaks. It is this conflict that Alex Wellerstein analyzes in Restricted Data: The History of Nuclear Secrecy in the United States, a comprehensive account of nearly a century of governmental efforts, often frustrated, to control information about nuclear technology.

Wellerstein, a historian of nuclear technology and a professor at Stevens Institute of Technology, has written extensively about the atom’s influence on our lives, publishing both in academic journals and in popular magazines like National Geographic; he has even released a video game, Oregon Road ‘83, about life in a post-nuclear United States. This body of work conveys one of the book’s best qualities: accessibility. Restricted Data is, by and large, written in plain language, carefully balancing readability and rigor to make this volume useful to both laypeople and experts.

This story begins with the advent of modern atomic theory. Leo Szilard (1898-1964), the Hungarian physicist at Princeton, inaugurated the concept of nuclear secrecy through a quixotic effort to persuade scientists to self-censor papers on nuclear fission, in hopes of preventing dangerous governments from understanding the military applications of fission. This period soon gave way to the Manhattan Project, and as the US government faced the project of maintaining secrecy about a program with thousands of participants, the classification apparatus grew larger and larger – and, notably, failed to prevent espionage, as the project was soon infiltrated by Soviet intelligence. Finally, he turns to the Cold War and, later, the unipolar world, all of which tended to expand secrecy over time even amid multiple declassification pushes.

Indeed, this book is in large part a record of historical failures. For one, Wellerstein fruitfully questions whether it is really possible to maintain secrecy around what is essentially physical law: any organization able to afford a nuclear program can ultimately unravel atomic “secrets” on its own, regardless of any secrecy policies. For another, he points out that, in general, the main obstacles to building a nuclear weapon are material, not informational; for example, the main bottleneck for the Soviet program was locating reserves of uranium, not in discovering how the bomb worked. For three, the government is often its own worst enemy: When the existence of an American hydrogen bomb project leaked, it wasn't espionage that did it: Senator Edwin Johnson simply blurted out the truth on television.

To make this case, Wellerstein relies primarily on documentary evidence. Of course, this is a challenging history to write, as so much of the source base remains classified. The author admits he has no security clearance, but having one would not help him – it’d simply subject his work to more government scrutiny. Nonetheless, he shows that much of the government’s deliberations about secrecy have been declassified, along with a wide variety of reprinted primary sources, ranging from diagrams of nuclear weapons to handwritten proposals for secrecy schemes. Taken together, these offer a great deal of insight into how policymakers wrangled with competing loyalties to national security, their own ideals, and democratic government.

It is this facet of the book that most demonstrates its value. It is hard for a scholar to be sympathetic to the censor, but Wellerstein urges us to resist seeing them as shadowy MIBs or corrupt conspirators. Rather, he points to figures like David Lilienthal, the AEC commissioner who entered the government as an idealist and left a broken, exhausted man, one of many would-be reformers “assimilated” by the nuclear establishment. Indeed, Wellerstein portrays censors here as deeply torn about their work, offering a nuanced portrayal of the price some pay when society agrees that some information should be kept secret, but offers no coherent schema for doing so.

Doubtless, Restricted Data is an invaluable contribution to its field, offering a comprehensive history of nuclear classification that is far more accessible than most scholarly texts. It deserves to be widely read – perhaps then the body politic would reflect more deeply on how much we want our government to hide from us.

Cite This Work