Spies, Lies, and Algorithms

This content has been archived. It may no longer be accurate or relevant.

From The Wall Street Journal:

Computers have transformed many institutions and professions in the 21st century, and the world of espionage especially. In “Spies, Lies, and Algorithms,” Amy Zegart, a Stanford professor of political science and an occasional consultant to intelligence agencies, has provided a lucid and sobering account of how digital and other technological breakthroughs are “generating new uncertainties and empowering new adversaries” for the United States at a time when its intelligence agencies are uniquely stressed.

Ms. Zegart opens her book with a survey of the nation’s rapidly changing “threat landscape” (Russia, China, terrorist groups); the sudden arrival of “open-source intelligence” (live-streaming amateur videos, time-stamped Twitter and Facebook posts); the consequently high volume of internet data relevant to intelligence; and the challenge to the U.S. intelligence community of keeping up with it all. Her aim is to give the general reader a non-Hollywood understanding of 21st-century intelligence as well as the daunting challenges that American spy agencies now confront.

The U.S. intelligence community, outlined concisely by Ms. Zegart, is composed of 18 separate organizations, including two independent agencies: the Office of the Director of National Intelligence, which oversees operations, and the Central Intelligence Agency, which runs spies and engages in covert action. There are nine Defense Department elements, including the National Security Agency, which makes and breaks code; the National Reconnaissance Office, which develops and deploys spy satellites; and the intelligence offices of the various armed forces. The other seven elements include divisions of the Department of Homeland Security, the Drug Enforcement Administration, the FBI and other government entities. Ms. Zegart catalogs the difficulties of coordinating these disparate organizations, each with its own culture and skill set and priorities.

As the world has become more and more connected electronically, so much data is now online—Ms. Zegart estimates that 80% of what the intelligence community gathers comes from publicly available sources—that intelligence agencies are losing their traditional advantages to nongovernmental actors. Spy satellites and high-resolution cameras mounted on military aircraft, once the exclusive preserve of the government, now have rivals in small commercial satellites that can observe even license plates from space. Costs for users have plummeted; Google Earth is free. A “cottage industry of non-governmental nuclear intelligence collectors and analysts” who track nuclear efforts in North Korea and Iran has emerged, along with such phenomena as the Netherlands-based Bellingcat, a private community of journalists and researchers that has provided remarkable information about the secret Russian unit that has attempted to assassinate dissidents in Europe.

While there are obvious benefits to such activities, the privatization of intelligence also has costs and dangers. Ms. Zegart is at her best when describing cyber threats. “In many ways,” she writes, “cyberspace is the ultimate cloak-and-dagger battleground, where nefarious actors employ deception, subterfuge, and advanced technology for theft, espionage, information warfare, and more.” Enemy states and terrorist groups are “hacking both machines and minds,” not only within American institutions but in our living rooms. “Artificial intelligence 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.”

. . . .

The anonymity of the internet combined with the widespread use of secure encryption has led to increasingly vitriolic and often hard-to-refute false claims clogging all channels of communication. Private technology companies like Google, Amazon and Facebook, incentivized to pursue profits and operating on a global scale, are often reluctant to surrender their users’ privacy even when democratic governments assert national-security claims. While China and other repressive regimes employ facial recognition software to harass, intimidate and imprison entire populations, many American Googlers have protested any cooperation by their employer with the American intelligence community.

The United States and other technologically advanced countries are increasingly vulnerable to large-scale cyberattacks that can corrupt data or compromise sensitive infrastructure. In 2015 a Chinese intrusion stole 21 million security-clearance records from the U.S. Office of Personnel Management; in 2020 Russia obtained access to several American nuclear labs, government departments and Fortune 500 companies.

Link to the rest at The Wall Street Journal (PG apologizes for the paywall, but hasn’t figured out a way around it.)