Renovating U.S. Intelligence: On Amy B. Zegart's "Spies, Lies, and Algorithms"
Amy B. Zegart | Spies, Lies, and Algorithms: The History and Future of American Intelligence | Princeton University Press | 2022 | 424 pages
In Spies, Lies, and Algorithms: The History and Future of American Intelligence, Amy Zegart, a political scientist from Stanford, proposes reshaping American intelligence organizations to meet the challenges of the twenty-first century. The United States has 18 intelligence organizations, each added in an ad hoc fashion following a major failure. Instead of inspiring awe or maximum efficacy, they underscore the limitations of the current structure of the United States intelligence community. The challenge facing the intelligence community and the U.S. as a whole now underlines the fundamental contradiction between business interests and national interests. Indeed, that challenge has little to do with the half-hearted coordination work between diverse and specialized agencies. Zegart contends that these two interests (business versus national) are irreconcilable, and unless some formula is found to harness business for the benefit of the nation, the operations of the intelligence organizations will stay largely dysfunctional and bypassed by tenacious and authoritarian adversaries.
Even with 18 intelligence organizations, the U.S. underperforms, in part, because satellite images made publicly available by Google have never ceased becoming better, complicating the work of intelligence analysts. The reason, Zegart thinks, is a lingering and counterproductive Cold War mindset. In an age of open-source information when the internet doubles its volume of information every two years, secrecy is increasingly difficult to maintain. Teenagers using Google Earth and other freely available and inexpensive applications can now access information that previously consumed considerable time and specially-trained personnel within government agencies. In this environment, where anyone can roleplay as a spy, and everyone with a reasonable set of skills can access sensitive data, secrecy is a liability, or perhaps a fiction. As such, the intelligence community needs to harness the courage to rethink its work.
To organize her proposal for revamping the intelligence community structure, Zegart lays out her argument slowly. In the chapter “Intelligence Challenges in the Digital Age: Cloaks, Daggers, and Tweets,” she handles the issue of power. She understands power more or less dialectically, as being a superpower translates to not only invincibility but also vulnerability. The second topic of focus is democratized data, which the internet revolution has introduced. Anyone with a computer and internet connection can monitor what Iran, North Korea, or any other government does not share. No state monopoly over access to sensitive information. This leads us to the third challenge, which is secrecy. In the past, maintaining secrecy used to give an advantage in intelligence collection. Now, secrecy is almost detrimental in the sense that no government can entirely protect its power grids, financial records, or start-up inventions—all of which can be accessed online—by disengaging from “the world.” Hence, Zegart’s argument that private actors such as Facebook, Amazon, and Google, among others, should be involved in securing America as most cutting-edge technologies can be used—and often are deployed—as weapons against American interests. This is a reason why secrecy in the old sense translates to disadvantages that severely hurt U.S. interests. The U.S. intelligence community has a lot of catching up to do when it comes to maintaining actual secrecy.
In Zegart’s chapter on what she calls “the education crisis,” she addresses what she calls the inhibitive impact of Hollywood. Spy entertainment—she calls it “spytainment”—provides a completely distorted image of intelligence work. Equally damaging for Zegart, spytainment clouds public perceptions of the real challenges facing America. Fiction maintains the myth that America is invulnerable, whereas, in reality, America is vulnerable. Besides, Hollywood fuels conspiracy theories such as former President Trump’s insistence on “deep state” plotting as an explanation for failed political maneuvers. Conspiracy has never been as devastating since congressmen and judges tend to believe spytainment with flat plotlines, featuring “heroes, escapism and the triumph of good over evil,” more than intelligence reports they have at hand. Clouded in secrecy, the culture of the supremacy of the intelligence agencies and a legacy of intelligence successes during the Cold War does not help face today’s challenges.
Zegart’s analysis of intelligence history provides a snapshot of the development of intelligence institutions since George Washington. She aims to remind policymakers and the general public alike that America is vulnerable. In its brief intelligence history, America could not overcome halted development, organizational fragmentation, and democratic tension. During the peace between World War I and World War II, the U.S. adopted a habit of dismantling its spy bodies.
As a latecomer to the spy industry, America should not be engrossed with its Cold War successes, particularly when compared to countries such as China, which has a millennial history of warfare and intelligence. The rules of the game are quickly changing and America—Zegart never tires of reminding—should not sleep on past feats. Zegart hammers the point that advances in technology are more disorienting than conducive to any clear strategic advantage. In her opinion, intelligence agencies should resist the temptation to violate their mission as information-gathering bodies. Their only task is to provide decision-makers an informational advantage.
One of Zegart’s strategies is to describe how intelligence operates in practice. She categorizes its three core missions as “the analytic,” “the human,” and “the operational,” all of which interact to make each intelligence organization what it is now. The analytic is geared toward giving policymakers an “advantage over adversaries.” For a successful execution of analytic missions, one has to be aware of the fine distinction between knowns and unknowns. We learn that intelligence should not be confused with the amassing of secrets, and as such, intelligence work has to be distinct from policymaking. The human side of the mission sheds light on a variety of motivations and traits that qualify typical analysts, officers, and informants. We read too about how intelligence officers balance their jobs with their private lives. There is a section on how officers grapple with moral dilemmas. In carrying out their mission, intelligence agencies handle interrogations of detainees. Still, evidence often amounts to no more than a good bet since cases where conclusive evidence can be reached are rare. Zegart finds that the golden rule with intelligence professionals is “…[the willingness] to challenge their prevailing hypotheses.”
Carrying out her argument for abandoning Cold War mindsets, “Why Analysis is so Hard: The Seven Deadly Biases,” Zegart underscores the gist of the book. Given the abundance of open-source data, the chapter seeks to answer why analysis has become excessively hard. Other than outside compromises, Zegart outlines the sinister role of seven deadly biases: confirmation bias, optimism bias, availability bias, fundamental attribution error, mirror imaging, framing biases and groupthink, and superforecasting. When an institution is sure it has neutralized dysfunctions such as “bureaucratic turf protection, agency cultures, career incentives, ingrained habits, and a desire for autonomy,” although not a simple task, it can move on to work on the seven biases. The key strategy to outsmart these biases lies in encouraging dissent; that is, finding a team of experts that reviews an intelligence case and advocates the devil’s view. She also notes that advances in artificial intelligence can help overcome human limitations.
Similarly, “Counterintelligence: To Catch a Spy,” grapples with traitors’ motivations, and how intelligence officers recruit informants in the digital age. Here we read that “China, Russia, Cuba, and Iran are among the most aggressive foreign intelligence services seeking to steal American secrets. Of them, China stands apart as the most serious counter-intelligence threat. American military experts have said that there isn’t a single major Chinese weapons system that isn’t based on stolen U.S. technology.” The chapter elaborates on early tell-tale signs for suspecting, investigating, and uncovering sell-outs (or moles in intelligence jargon) without compromising trust among members of the intelligence community. Three key counter-intelligence challenges are trusting too much; paranoia, or trusting too little; and technology that makes it possible to recruit assets from afar. Technology makes it equally possible to incur considerable damage if a trusted insider breaches his or her trust oath. For example, we read how the damage done by turncoats such as Snowden has been irreparable.
Likewise, “Covert Action: A Hard Business of Agonizing Choices” is where Zegart’s studies undercover operations serving a certain line of policy but which can either be claimed or officially disowned depending on interest, not on success or failure. The operation that killed Bin Laden in 2011 counts as one, as does the CIA’s funneling of money to help Italy’s Christian Democratic Party to win parliamentary elections in 1947. Since only the president can authorize covert actions, the chapter weighs those uneasy choices presidents take or circumvent to serve a policy. When all policy lines have been tried and exhausted, covert actions serve as the last resort. The ways drone technology and the war on terror have factored into recent engagements force policymakers to face how the blurring of intelligence and military mandates is counterproductive.
But democracy itself can be a liability, as Zegart specifies in “Congressional Oversight: Eyes on Spies.” She recounts that as lawmakers, congressmen are not trained or sufficiently motivated to do the oversight work stipulated by the constitution. Zegart summarizes the challenges facing congressional intelligence committees in three words: information, incentives, and institutions. Given the inhibitive influence of spytainment in addition to the poor payoff from carrying out proper oversight of intelligence organizations, Zegart observes an informational and motivational lag beneath successive congressional committees charged with cross-checking intelligence organizations. She highlights a deeper structural problem of these committees’ culture that does not encourage rigorous second opinions when it comes to the details of intelligence agencies’ daily work. The compounding effect of these three challenges explains scandals such as the presumed weapons of mass destruction owned by Iraq. One comes face-to-face with just how quickly policy is outpaced by technology.
For purposes of drawing a convincing point against the secrecy mindset, “Intelligence Isn’t Just for Governments Anymore: Nuclear Sleuthing in a Google Earth World,” Zegart further advances the cause of renovating U.S. intelligence. Underneath the chapter lies a call for humility as “estimating nuclear threats is hard. Assessing the intelligence track record is, too.” The democratization of intelligence is a relatively new phenomenon that breaks down a government’s monopoly over sensitive information. Low-cost commercial satellites with imaging capability that nearly equals military satellites are routinely put in orbit. Machine learning and computer modeling are enhancing the identification of surface-to-air missile launching sites for anyone with an internet connection and the patience for tracking terrestrial alterations. Hobbyists using only Google Earth images can track Iran or North Korea’s uranium-enrichment facilities and the activities taking place therein. Once the intelligence ecosystem is widely open to non-governmental actors, intelligence policy has to accommodate the informal branch lest foreign intelligence services enlist this growing informal branch.
Putting those policy makers and her readers on the same page with the real, not imagined, dangers of the cyber age, Zegart, in “Decoding Cyber Threats,” elucidates how the internet has opened the door for a new generation of warfare rooted in deception, sabotage, and misinformation. Hacking and deepfakes can sow the seeds of social discord and upheaval. The examples that Zegart uses to illustrate her point are telling: Shadowy Kremlin-backed organizations armed with automated Facebook accounts or bots sow discord in American cities. The 2016 presidential elections are registered by the intelligence community as a cyber Pearl Harbor. We read too that “China is believed to have stolen trillions of dollars of intellectual property, including terabytes of data and schematics for the F-35 and F-22 stealth fighter jet programs.” Without cooperation between the private sector and state agencies, such complex intelligence challenges triggered by the digital age cannot be overcome—and the cost will be American democracy and liberalism. This explains Zegart’s initial call to rethink the structuring of intelligence agencies along lines that do not abandon Cold War methods but consider the need to engage with open-source data and other unorthodox initiatives.
The book pulls thirty years of research and experience advising the U.S. government together with hundreds of interviews with current and former intelligence people. As a career academic, Zegart comes to these agencies as an outsider—a position which works to her advantage since probably only an outsider can reflect honestly on factors which weaken the institutions’ ability to effectively address new threats.
Contrary to Hollywood’s overblown portrayals of American invincibility, the historical record of American intelligence agencies, though professional and functional, is far from adequate in meeting cyber threats and other security challenges of the digital age. What Zegart has in mind is the recent failure of America’s spy network in China. Hence the call for renovation and accommodating a new digital reality is nothing short of a call for revolution. Zegart states that “today’s technological demands, though, are even greater because there are more breakthrough technologies. They’re spreading faster and further. They’re inherently hard to understand. They’re driven by commercial companies seeking global markets, not governments seeking national security.” Here we realize that Zegart has touched on the core of the problem without registering exactly the full implication that the problem propagates. America is experiencing a self-contradiction in movement: the ideals of nationalism versus the calls of globalization. It is time for the American establishment to decide whether they should go for “American capitalism” or borderless capitalism, a truly global system.
All else, such as debates over the competency of congressional oversight, cyber threats, and breaches of secrecy, are secondary once the earlier question is resolved. Addressing the efficiency of democratic measures in the form of congressional oversight to prevent personal or institutional abuses becomes a liability, a crippling structure. Because authoritarian regimes are free from democratic stipulations in their accountability system, they have an advantage over America.
Indeed, it is not the lack of patriotism and sense of national service among those heading tech companies that drives the present fixation on U.S. intelligence. Predisposed to markets, tech companies’ allegiance is to shareholders, not the general public. To account for this contradiction, Zegart implicitly suggests that democracy should make some sacrifices “because the sources of dysfunction run deep—in information, incentive, and institution.”
Other than being an discreet call for jingoism, the problem with the book is that it sees intelligence organizations and the state that these organizations presumably protect as independent entities. It is useful to recall that the successes of World War II and the Cold War were made possible by economic miracles as U.S. companies, not the U.S. government or its bodies, beat all competitors combined, foes and allies alike. Today’s corporate hunt for profit presupposes that any allegiance to the state leads to reduced competition. In economic parlance, reduced competition is synonymous with asphyxiation. Between asphyxiation and global growth, tech companies have chosen the latter. Thus, the state with its 18 intelligence bodies can do very little except postpone, not reverse, the collapse of the Westphalian state order. Instead of addressing the major transformation ahead, Zegart chooses to ruminate on how companies should be loyal.