Spies, Lies, and Algorithms

The History and Future of American Intelligence


Amy B. Zegart, Princeton University Press, Princeton, New Jersey, 2022, 424 pages


Download the PDF Download the PDF

Lt. Col. John H. Modinger, PhD, U.S. Air Force, Retired, Fort Leavenworth, Kansas

Book Cover

In this excellent exploration of the challenges facing intelligence agencies in the decades to come, academic and policy wonk Amy B. Zegart welcomes the uninitiated reader into this frenzied world of change, automation, and deception in the book Spies, Lies, and Algorithms: The History and Future of American Intelligence.1 Her status as an academic versus an operator has, she acknowledges, both drawbacks and benefits. While she does not have access to classified records, she claims an outsider’s perspective, complete with ample doses of skepticism and independence, which leaves her free to pose uncomfortable questions and arrive at unflattering conclusions. The implication is—obviously—more so than an insider.

The backdrop against which intelligence services will have to operate is, in many respects, a blur. It is expected that artificial intelligence may possibly eliminate 40 percent of all jobs in the next twenty-five years. “Not since electricity has a breakthrough technology ushered in so much potential promise and peril.”2 In 2019, Google announced it had achieved “quantum supremacy”—a computing breakthrough so powerful that a math problem a supercomputer would need ten thousand years to solve could be cracked by its machine in just three minutes and twenty seconds. That kind of computing power could eventually unlock the encryption protecting nearly all the world’s data today, placing at risk the entire financial system. Similar advances are underway in synthetic biology and nanotechnology as well.

Of course, we have seen technological advances before. But what sets this time apart from the past is the convergence of so many new technologies changing so much so fast. Zegart identifies three profound ways in which tech is challenging the intelligence-gathering landscape. First, the pace of change is generating new uncertainties and enabling new adversaries. While the Cold War was a dangerous time, it was, in many respects, a simpler time, too, and America’s intelligence priority was clear. Today, by contrast, an array of bad actors is leveraging technology. A second challenge is the data itself. Intelligence is a sense-making enterprise. It used to be that a small contingent of spy agencies dominated the collection and analysis of information. Now, data is democratized, and the “old pros” are struggling to keep up. The sheer volume of data is staggering, let alone trying to comprehend it all. It is estimated that the amount of information on Earth is now doubling every two years, and spy agencies are drowning in data. The intelligence playing field is leveling—and not in a good way.3 In this topsy-turvy world, intelligence agencies are struggling to adapt. Suffice it to say, whoever can leverage all this data better and faster will win. A third challenge is secrecy. There has always been a built-in tension balancing secrecy and openness. Secrecy is essential to protect sources and collection methods; openness is vital for securing democratic accountability. Too much secrecy invites abuse; too much transparency renders intelligence unproductive. In the “good ole days,” tech advances (like the internet) usually started inside the government and then migrated to the commercial sector. Today, that process is reversed. Instead of developing tech in-house, spy agencies now have to spot them and adopt them rapidly from outside entities. And the gravitational pull of these companies is powerful. It is difficult for government agencies to compete with companies that offer lucrative salaries and cutting-edge computer facilities. Then there is the persistent problem of engagement and collaboration with these companies. The level of distrust vis-à-vis spy agencies is high, and the history is filled with many dark chapters. Those agencies are attempting to win over skeptical companies and rebuild trust, but it will not be forged easily … or quickly. Another thing that is changing is who is making the decisions. Once upon a time, important decisions were made by federal agents who wore badges, held high-level security clearances, and knew how the intelligence community (IC) worked. Today, decision-makers often live a world apart from Washington and are more focused on quarterly profits and shareholder interests than they are national security.

Zegart’s purpose in writing this tome is to offer readers a better understanding of the intelligence world and the challenges intelligence agencies now face. She readily admits there are no easy answers but insists one thing is definitely true: America’s intelligence agencies must adapt, or they will surely fail.

The book covers the gamut from intelligence challenges in the digital age to the mindboggling extent to which fictional spy depictions are shaping public opinion and intelligence policy. From American historical spying at a glance to some knowns and unknowns about the business, it is a deep look into why the work of analysis is so hard given the biases always lurking in the minds of analysts—cognitive traps that can lead even the sharpest minds astray. It delves into topics ranging from the tough work of counterintelligence to covert action and the hard business of tough choices—what covert action is and why all presidents use it, even though it so often fails. It examines the world of congressional oversight—from why it matters, to why it often fails to achieve its goals, and what the future holds—and reviews some highly contentious detention, interrogation, and warrantless wiretap programs. It also looks at the other players in the intelligence game these days. Finally, it includes a sober, albeit depressing, treatment of the cyber threats out there.4 As Zegart puts it, “AI is creating deepfake video, audio, and photographs so real, their inauthenticity may be impossible to detect. No set of threats has changed so fast and demanded so much from intelligence.”5

For America’s IC, the new digital age is bursting with complexity and challenges. Success in this brave new world will demand a fundamental rethink about how to secure advantage in a radically new landscape. For Zegart, that rethink begins by getting back to basics and depoliticizing intelligence. But it will require more than just that. Agencies will have to embrace open-source intelligence, develop new capabilities, and reward agents and actors for doing things differently. Undoubtedly, this involves a tremendous paradigm shift, but one she argues is essential.

Over the last couple of decades, spy-themed entertainment has exploded while actual spy facts remain cloaked in darkness. Maybe unsurprisingly then, “spytainment” is standing in as a substitute for adult education on the subject. While this may seem laughable and exaggerated, there are increasingly far-fetched instances where fictional spy tales are shaping public opinion and, more concerning, intelligence policy. Today, Hollywood is releasing twice as many spy blockbusters as it did in the 1980s.6 That trend, toward more and more spy thrillers, seems likely to ensure distortions between reality and fiction will only grow in the years ahead. Indicative of this phenomenon was the box office hit Zero Dark Thirty, the Academy Award-nominated film about the Central Intelligence Agency’s (CIA) ten-year hunt for Osama bin Laden. The filmmakers received considerable help from the CIA crafting the story. Maybe unsurprisingly, the film portrays the agency in a flattering light. However, when the film was released, it ignited so much controversy about what was real and what was not that the acting CIA director, Michael Morell, had to issue a memo to his own workforce clarifying the facts. “What I want you to know is that Zero Dark Thirty is a dramatization, not a realistic portrayal of the facts.” Morell felt he could not stop there. He clearly felt compelled to directly dispute the movie’s central claim: that harsh interrogation methods were pivotal to tracking bin Laden to his compound in Pakistan. “The film creates the strong impression that the enhanced interrogation techniques that were part of our former detention and interrogation program were the key to finding Bin Laden. That impression is false. The truth is that multiple streams of intelligence led CIA to conclude that Bin Laden was hiding in Abbottabad.”7 “Reality is nuanced. The movie was not. The result was deeply misleading. [T]oo often, depiction is shoddy education.”8 While spy fiction has become widespread, actual spy facts remain elusive. The net effect is a confusing maze where it is difficult and time-consuming to untangle what is real and what is pretend.

Simultaneously, many officials, together with critics of certain intelligence agencies, frequently complain that far too much information is needlessly classified, impeding information sharing, hindering collaboration, and undermining democratic accountability. Despite their complaints, little has changed in the many decades since the complaints began because bureaucracies naturally hoard information since revealing secrets can get bureaucrats into trouble but keeping them rarely does. The prevailing, but infuriating, mantra continues to be, “When in doubt, classify.” There is frustration, even at the highest levels. Donald Rumsfeld’s own deputy secretary of defense for counterintelligence and security testified before Congress that “half of all classification decisions are unnecessary over-classifications.”9 More recently, Gen. John Hyten, vice chairman of the Joint Chiefs of Staff, declared, “In many cases … we’re just so overclassified it’s ridiculous, just unbelievably ridiculous.”10

Another barrier to understanding intelligence is, no doubt, the honed culture of secrecy embedded in the IC itself. That culture has a powerful grip. While a great deal of secrecy is required for intelligence folks to do their jobs, much of it is overkill. Zegart relates how she came to learn of a strategic plan which warned about terrorism and called for radical Federal Bureau of Investigation reforms three years before 9/11. Unfortunately, those reforms were never implemented. It’s not difficult to understand why an agency would prefer not to have that kind of information out there. It would embarrass many and put the agency in a bad light. Another facet of this wall cordoning off intelligence, making it hard to comprehend, is that few study or teach it, largely because of the dearth of data. Data and evidence are the lifeblood of academic research. If one wants to study congressional decisions, there is a trove of data; legal scholars, likewise, have a bounty of court cases and judicial opinions at their fingertips. By contrast, intelligence scholars like Zegart have to rely on whatever gets declassified or whatever declassified information they can assemble. And while there is the Freedom of Information Act, the responsibility of government agencies to respond is not synonymous with handing over documents.11 In short, it can be a Herculean task to get the information, even if the act allows for it. It is hardly surprising, then, that many professors avoid intelligence research. With ticking tenure clocks, they are often unwilling to risk their futures betting on the accessibility of documents. Thus, and regrettably, “spytainment” is all many have to get familiar with the IC.

But the proliferation of spytainment creates some real problems. First, it creates a public mindset that views intelligence agencies as far more powerful, capable, and unaccountable than they actually are; many come to believe these agencies are omnipotent, fueling, among other things, “Deep state” conspiracy thinking, which purports the government is out there running rogue, and presumably acting against most people’s interests on behalf of some murky, ill-defined power.12 But that kind of wrongheaded thinking is not limited to uneducated or ignorant masses. No, policymaking elites are, all too often, invoking fictional spies and unrealistic scenarios as they formulate actual intelligence policy. So, “from the heartland to the beltway, a little knowledge of intelligence turns out to be a dangerous thing.”13

Of course, the internet has facilitated “an online ecosystem tailor-made for spreading false narratives at lightning speeds and unprecedented scale. The internet has become a misinformation superhighway where conspiracy theories can be conjured up by anyone, posted on social media, spread by hashtag, amplified by bots, and picked up by mainstream media—all at the touch of a button.”14 In this “wild, wild West” forum, conspiracy theories can—and are—peddled by millions, including even one former president. The Trump administration went to great lengths to accuse U.S. intelligence agencies of having a role in a so-called deep state that was supposedly and secretly working to undermine his presidency at every turn. He infamously said he trusted Russian President Vladimir Putin’s denials of interfering in the 2016 presidential election more than the judgments of his own intelligence agencies. This was/is astonishing!15 The troubling allure of conspiracy theories and deep state mythology raises serious doubts about the ability of intelligence agencies to fulfill their mission in the future, if such large portions of the public—and even a president—view them with such disdain and distrust.

So, while “spytainment” is engrossing and fun to watch, it has a distinct downside. Too often, fiction supplants fact, creating fertile ground for conspiracy theories to grow and shaping the formulation of actual intelligence policy. In the end, secret agencies in democratic societies cannot succeed without trust from those it serves. But trust requires knowledge. As former CIA and National Security Agency director Gen. Michael Hayden plainly puts it, “The American people have to trust us and in order to trust us they have to know about us.”16 How one goes about accomplishing that while preserving operational effectiveness remains to be seen, but he is certainly right.

There are so many appealing aspects of this book that this reviewer is inclined to wax on. But to go through each chapter at length would make this essay too unwieldy, too wordy. It is hoped this appraisal sheds sufficient light on Zegart’s home run that readers are encouraged to grab a copy and indulge their curiosity. Suffice it to say this is an outstanding bit of scholarship a reader will have a hard time putting down given the treasures awaiting one’s eyes on each successive page. If you want to thoroughly understand the panoply of challenges facing the intelligence community today, this book is a must read and comes with my highest recommendation! You will not be disappointed.


  1. Amy Zegart served on the National Security Council staff and advised policymakers but never worked inside an intelligence agency. Instead, she is a career academic who has examined spy agencies from the outside, looking at how they have evolved over time, why they struggle so mightily to adapt to new threats, and what they must do to improve.
  2. Amy B. Zegart, Spies, Lies, and Algorithms: The History and Future of American Intelligence (Princeton, NJ: Princeton University Press, 2022), 2.
  3. Ibid., 6.
  4. “Spytainment” has fueled conspiracy theories and influenced policymakers, with eighteenth-century invisible ink to twenty-first-century spy satellites; what intelligence is, what it is not, and how it operates. What motivates trusted insiders to become traitors?
  5. Zegart, Spies, Lies, and Algorithms, 15.
  6. Ibid., 25.
  7. Ibid., 27.
  8. Ibid., 28.
  9. Ibid., 30.
  10. Ibid.
  11. Freedom of Information Act, 5 U.S.C. § 552 (1967).
  12. Zegart, Spies, Lies, and Algorithms, 37. According to a 2006 Scripps poll, 36 percent of Americans considered it “likely” or “somewhat likely” that U.S. government officials either carried out the 9/11 attacks or knowingly allowed them to occur. Ten years later, another poll found that 25 percent of Americans still believed it was “probably” or “definitely” true that the “U.S. government helped plan the attacks of 9/11” despite no evidence that it is true and overwhelming evidence that it is not.
  13. Ibid., 36.
  14. Ibid., 37.
  15. Ibid., 38. A 2017 Washington Post–ABC News poll found that nearly half of all Americans (48 percent) believed that a deep state of “military, intelligence and government officials who try to secretly manipulate government policy” exists; only 35 percent thought the deep state was an unfounded conspiracy theory (which means 65 percent thought the deep state existed in one form or another).
  16. Ibid., 43.


Lt. Col. John H. Modinger, PhD, U.S. Air Force, retired, is an associate professor with the Department of Joint, Interagency, and Multinational Operations at the U.S. Army Command and General Staff College, Fort Leavenworth, Kansas.


Military Review

Book Review Program

The Military Review book review program allows reviewers to read books of interest to military professionals—often before book publication—and then present their thoughts on the Army University Press website. The reviewer then retains the book. Read our latest book reviews at http://www.armyupress.army.mil/Journals/Military-Review/MR-Book-Reviews/.

Books for review are available only through the Military Review book review editor. If you are interested in becoming a reviewer, see our Book Review Submission Guide at http://www.armyupress.army.mil/Journals/Military-Review/MR-Book-Review-Submission-Guide/.



Back to Top

March-April 2023