In this edition of Author Talks, McKinsey Global Publishing’s Raju Narisetti chats with academic and political scientist Amy Zegart about her new book, Spies, Lies, and Algorithms: The History and Future of American Intelligence (Princeton University Press, February 2022). Zegart is a senior fellow at the Hoover Institution and the Freeman Spogli Institute at Stanford University, where she is also a professor of political science.
With the proliferation of advanced technologies like AI and the internet, the amount of data in the world is estimated to double every two years. Yet, Zegart says the public’s understanding of intelligence gathering is minimal. An edited version of the conversation follows.
Why is a book on US intelligence timely?
This moment is particularly relevant for thinking about challenges to US intelligence because we’re at a technological juncture. Never before have we faced so many emerging technologies, from AI to quantum computing to synthetic biology.
These emerging technologies are changing not only the business world but also the world of espionage. It’s an adapt-or-fail moment for US intelligence agencies.
Why is the proliferation of ‘spytainment’ a problem?
I found the rise of spy-themed entertainment has had some unintended and pernicious consequences for both public opinion and public understanding of intelligence and for policy makers and how they reason through real intelligence challenges.
Just to give a few examples, I found in polls of my students that they actually didn’t know much about intelligence. What they learned from spy-themed entertainment was wrong. I also found that policy makers were using fictional plotlines from shows like 24 in Senate confirmation hearings for a CIA director, and Supreme Court justices were using fictional plotlines in reasoning how they would decide on certain cases—so, spy-themed entertainment isn’t just entertainment.
Why does AI not always make for better intelligence gathering?
New technologies like AI can make intelligence better and make analysis more effective and time efficient, but the intelligence community has to adopt those technologies to be able to use them. That’s much harder than it might seem.
In the Cold War, new innovations in technology started in the government and then became commercialized, like GPS or the internet, for example. Today, it’s the opposite; new technologies like the latest AI come from the private sector first, and the government has not been as adept at adopting those technologies at scale as it should be.
How does the government’s data collection compare with that of the private sector?
The National Security Agency [NSA] has a lot less data about me than, say, Google does. Google knows where I shop, what I like to buy, how I spend my time on the internet, what sites I visit, and where my information comes from.
That’s information the US government does not have, and Google can use that data—sell it or use it—in ways that I may not understand or may not even know about. Private-sector companies have a lot more information about the day-to-day activities of citizens than the government does.
Tech companies face a real challenging moment. They have global shareholders, global markets, and global employees, but they’re also caught in the crosshairs of geopolitical competition where economics and security are intertwined.
Technology isn’t just for commercial use. New technologies are inherently dual use, and they also have military applications. This is a new era where there has to be more collaboration, more discussion, and more partnership between tech companies and the US government—and that’s challenging. Their incentives are not perfectly aligned, and it’s a work in progress.
Why should the US embrace open-source intelligence?
Open-source intelligence is any intelligence that is publicly available: Think about tweets on Twitter, or things that are posted online, or publicly available information in paper form that governments or organizations provide, like newspapers.
What’s happened now—in the internet era—is the amount of open-source intelligence is astronomical, and it’s growing exponentially. By one estimate, the amount of data on Earth is doubling about every two years. This means that open-source intelligence can provide a treasure trove of information—not just for government spy agencies but for anybody with an internet connection or a cell phone. Spying isn’t just for governments anymore. There’s lots more competition for the collection and analysis of information.
What’s happened now—in the internet era—is the amount of open-source intelligence is astronomical, and it’s growing exponentially. By one estimate, the amount of data on Earth is doubling about every two years. This means that open-source intelligence can provide a treasure trove of information—not just for government spy agencies but for anybody with an internet connection or a cell phone.
How much is the topic of intelligence researched and taught?
One of the things I found in doing research for this book is that most universities—at least the universities ranked in the top 25 by U.S. News & World Report—offer courses on the history of rock and roll rather than US intelligence. So, I joke to my students that they have a better chance of learning about U2, the band, rather than U2, the spy plane.
I also found that political-science professors like me don’t spend a lot of effort researching intelligence agencies. I looked at the top political-science journals over a 15-year period, and they published a total of 2,700 articles.
Of those articles, only five of them covered any topics even remotely related to US intelligence. So, there’s a disconnect at the exact moment that intelligence challenges and controversies are front and center in the news and front and center for policy makers. Political scientists have been studying just about everything else.
How is mainstream media coverage of intelligence?
One of the things I found with my national polling is that mainstream media news coverage is not enough. In 2013, for example, when former NSA contractor Edward Snowden stole and released an enormous quantity of highly classified documents exposing NSA surveillance programs, the news was awash with headlines about Snowden and what he revealed.
I polled Americans. I fielded a national poll a few months after those revelations amid this media frenzy, and what I found was most Americans had no idea what the NSA even did.
Only half of Americans actually correctly identified one of the NSA’s core missions: making and breaking codes. The large majority of Americans, 75 percent or more, either didn’t know or wrongly thought that the NSA did things like interrogate terrorist detainees. That suggests that all of this media coverage is happening against a backdrop of ignorance or misconception about what our intelligence agencies do, and the coverage isn’t breaking through.
Only half of Americans actually correctly identified one of the NSA’s core missions: making and breaking codes. The large majority of Americans, 75 percent or more, either didn’t know or wrongly thought that the NSA did things like interrogate terrorist detainees. That suggests that all of this media coverage is happening against a backdrop of ignorance or misconception about what our intelligence agencies do, and the coverage isn’t breaking through.
What was your biggest surprise in researching this book?
I originally started out writing a different book than the book that I wrote. I was originally going to write what I affectionately referred to as “Intelligence 101”—sort of how the world of intelligence operates. What surprised me most was how I ended up writing about “Intelligence 2.0,” which is how emerging technologies are challenging US intelligence agencies today.
They’re doing it in five ways. I call them the five “mores.” The convergence of technology is creating, number one, more threats for the United States—more threats through cyberspace, in particular, that our intelligence agencies need to understand. The second more is more speed: the acceleration of decision-making time means that intelligence has to operate at the speed of networks, not the speed of bureaucracy.
The third more is more data; intelligence analysts, like the rest of us, are drowning in data. The fourth more is more consumers—more decision makers outside the government who need intelligence. Think about voters who need intelligence about foreign election interference or tech leaders who need intelligence about cyberthreats.
Then there’s the fifth more: more intelligence competitors. I devoted two years and a whole chapter to “nuclear citizen detectives” who are tracking the most secretive nuclear threats around the globe using only unclassified and publicly available information like commercial satellite imagery.
What does a smarter intelligence apparatus look like?
Success is so hard to gauge in the intelligence business because, often, success is the nonoccurrence of events. We’ll never know what we stopped from happening because of intelligence success.
If I look at process metrics of success, future success lies in the fact that our intelligence agencies are harnessing this open-source intelligence world; that they’re interacting and engaging with open-source intelligence organizations and that ecosystem; and that they’re providing what we call decision advantage to policy makers—that policy makers feel their intelligence agencies are supporting them fast enough and fully enough so they can see threats and opportunities coming.
What can you tell us about your McKinsey years?
I’m a proud McKinsey alum. I have to say I find my experience at McKinsey—first as an analyst and then as a consultant—is the gift that keeps on giving. You know, in academia, we often joke, “That idea seems to work in practice, but the real question is does it work in theory?”
My McKinsey training really grounded me in marrying the idea of theory and practice and thinking about how organizations can adapt to new environmental challenges. At McKinsey, I did that in the business world, and I’ve taken that general approach to my research about US intelligence agencies: How can they adapt better to serve the country in a changing threat landscape?
Watch the full interview
