In an age of artificial intelligence, satellite surveillance, and signals interception that can vacuum up billions of communications, human intelligence should be obsolete. The logic seems obvious: why risk a spy when algorithms can do the work? Yet a century after the first modern intelligence agencies took shape, human intelligence remains not just relevant but irreplaceable. The technology that was supposed to kill it has instead forced it to evolve.
Human Intelligence in the Cold War Crucible
The foundations of modern human intelligence were laid during the Cold War, when the CIA and KGB refined techniques that would define tradecraftThe specialized techniques and methods used by spies and intelligence operatives, including codes, disguises, and covert communication. for generations[s]. Both superpowers employed spies, defectors, and informants to infiltrate enemy territory, developing an elaborate vocabulary of dead dropsA delivery method where vendors hide packages at specific GPS coordinates for buyers to retrieve, minimizing direct contact between parties., brush passesA covert tradecraft technique where two operatives briefly exchange materials in a public setting without appearing to interact., and surveillance detection routesA planned route an intelligence operative walks before meeting an asset, designed to reveal whether they are being followed..
Few cases illustrate Cold War human intelligence better than Adolf Tolkachev, a Soviet radar engineer who volunteered to spy for the CIA in 1977. Tolkachev approached American officials multiple times over two years before the Agency finally agreed to meet him[s]. The caution was warranted: the KGB routinely dangled fake volunteers to expose CIA methods. But Tolkachev was genuine, and over the next seven years he passed secrets on Soviet avionics, cruise missilesA guided missile that flies at low altitude using onboard navigation to reach its target with high precision, as opposed to a ballistic missile., and radar technology that saved the United States an estimated $2 billion in weapons research[s].
The CIA recruited Tolkachev using methods that would be recognizable to any case officer today: the SADRAT cycleThe CIA's six-step agent recruitment process: Spotting, Assessing, Developing, Recruiting, Agent handling, and Termination of a human intelligence source. of spotting, assessing, developing, recruiting, agent handling, and termination[s]. These fundamentals have proven remarkably durable. Russian intelligence services, drawing on over a century of experience from the Cheka to the KGB to today’s SVR and GRU, still rely on compartmentalizationA security principle that limits access to information by dividing it into separate, need-to-know segments to prevent unauthorized disclosure., deniability, patience, and psychological control[s].
The Digital Surveillance Threat
The 21st century brought challenges that Cold War spymasters never imagined. China now operates an estimated 540 million surveillance cameras, with Beijing claiming total CCTV coverage of the capital[s]. In 2017, China began fingerprinting all foreigners entering the country, making alias travel nearly impossible[s].
The 2010 assassination of Hamas commander Mahmoud Al-Mabhouh in Dubai demonstrated the vulnerability. Basic CCTV cameras across luxury hotels allowed investigators to unravel the movements and disguises of the entire Israeli team[s]. Biometrics now tie biological characteristics to specific identities, limiting intelligence officers’ ability to move under alias[s].
Digital footprints compound the problem. Every device interaction creates traceable records. Cover identities now require years of backstopped social media presence to withstand scrutiny. The old world of human intelligence, where a forged passport and a good legend could carry an operative across borders, has fundamentally changed.
The Counter-Terror Renaissance
Paradoxically, the September 11, 2001 attacks revealed that human intelligence remained essential precisely when technology dominated. The War on Terror necessitated renewed emphasis on infiltrating terrorist organizations and thwarting attacks[s]. According to former CIA Director John Brennan, the United States misjudged Saddam Hussein’s weapons capabilities in 2003 largely because it lacked human intelligence assets in Iraq[s].
Human intelligence may comprise only 10 to 20 percent of intelligence analysis inputs, with technical collection providing the rest[s]. But those human sources often provide precisely what technology cannot: adversary intentions. Understanding what Vladimir Putin or Kim Jong Un plans to do next requires human access, not satellite imagery.
When Cyber Needs Humans
The Stuxnet operation against Iran’s nuclear program, discovered in 2010, became the defining example of cyber-physical fusion. The malware was engineered to destroy centrifuges at Natanz, a facility air-gapped from the internet. Technology alone could not reach it. Intelligence sources widely suggest Stuxnet entered Natanz via an infected USB drive, carried into the highly secure facility by an unsuspecting contractor or insider[s].
The worm destroyed approximately 1,000 of Iran’s 5,000 centrifuges and delayed the nuclear program by an estimated two years[s]. Lines of code achieved what bombs might not have, but human intelligence delivered the payload. This remains the pattern: hostile actors use insiders or access agents to provide unique types of access impossible by computer network exploitation methods alone[s].
Tradecraft Adapts
Operation Ghost Stories, the FBI’s ten-year investigation culminating in 2010, exposed a Russian spy ring operating across the United States[s]. The illegalsIntelligence operatives deployed abroad under false identities without diplomatic cover, operating as ordinary citizens while conducting espionage. never obtained classified documents, but they were playing the long game: developing sources in policymaking circles who might one day hold power[s]. It echoed the Cambridge Five, Soviet talent spotters who recruited Cambridge University students in the 1930s and waited decades for them to rise through British intelligence[s].
Russian tradecraft has adapted to technology while preserving fundamentals. Dead drops, what Russians call “tainiki,” remain standard: magnets under bridges, hollowed-out stones, waterproof capsules in parks[s]. Communications have evolved to include encrypted apps and steganographyThe practice of hiding secret messages within ordinary text or images so the existence of the message itself is concealed. in digital images, but the underlying logic persists. The KGB used markings on telephone and utility poles to signal John Walker during his 1980s betrayal of the U.S. Navy[s]. Some Russian operations still do.
Human Intelligence and the Future
In 2021, MI6 Chief Richard Moore gave a rare public speech acknowledging the paradox facing modern intelligence agencies. “To stay secret, we are going to have to become more open,” he said[s]. Moore noted that technological progress in the next decade might equal that of the last century[s]. But he also emphasized the enduring necessity of human sources: “Even in a digital world, critical decisions are made by real people”[s].
The United States now spends $72.4 billion annually on intelligence, funding 18 agencies and roughly 100,000 employees[s]. Much of that goes to technical collection: satellites, signals intercepts, cyber operations. Yet human intelligence persists because no amount of technical prowess can replace a well-placed source who knows what an adversary intends.
The spy trade has survived by doing what it has always done: adapting. Cover identities now require years of digital backstopping. Biometric collection has made alias travel harder but not impossible. The fundamentals, recruiting human sources willing to betray secrets, remain unchanged since the Pharaohs deployed agents to gather intelligence on enemies. The brush of a hand against a park bench, a hollow rock in a quiet corner, the cultivated relationship that yields information no satellite can see: these ancient techniques endure because technology still cannot replicate trust, motivation, or human judgment. Human intelligence survived the digital age by becoming part of it.
In an age of artificial intelligence, satellite surveillance, and signals interception capable of vacuuming billions of communications, human intelligence should be obsolete. The logic seems straightforward: why risk a spy when algorithms can do the work? Yet a century after the first modern intelligence agencies took shape, human intelligence remains not just relevant but irreplaceable. The technology that was supposed to kill it has instead forced it to evolve, adapt, and ultimately prove its enduring necessity.
The Cold War Crucible of Human Intelligence
The foundations of modern human intelligence were forged during the Cold War, when the CIA and KGB refined techniques that would define tradecraftThe specialized techniques and methods used by spies and intelligence operatives, including codes, disguises, and covert communication. for generations[s]. Both superpowers employed spies, defectors, and informants to penetrate enemy territory, developing an elaborate operational vocabulary: dead dropsA delivery method where vendors hide packages at specific GPS coordinates for buyers to retrieve, minimizing direct contact between parties. (impersonal exchanges), brush passesA covert tradecraft technique where two operatives briefly exchange materials in a public setting without appearing to interact. (quick handoffs in public), SDRs (surveillance detection routesA planned route an intelligence operative walks before meeting an asset, designed to reveal whether they are being followed.), and one-time padsAn encryption method using a random key the same length as the message, used only once, making it theoretically unbreakable when properly applied. for unbreakable encryption.
Few operations illustrate Cold War human intelligence tradecraft better than the Tolkachev case. Adolf Tolkachev, a Soviet radar engineer at the Phazotron design bureau, first approached American officials at a Moscow gas station in January 1977. The CIA initially refused contact, fearing a KGB dangle operation. Tolkachev persisted for two years, making multiple approaches and eventually providing detailed technical intelligence to prove his bona fides[s].
Once recruited, Tolkachev photographed classified documents in office bathrooms where lighting was better and surveillance lighter. He communicated via dead drops using operational securityPractices that protect sensitive activities from adversaries by controlling what data is revealed about one's identity, location, or methods. materials including secret writing carbons, one-time pads, and pre-written cover letters. Over 21 clandestine meetings spanning seven years, he passed intelligence on Soviet avionics, cruise missilesA guided missile that flies at low altitude using onboard navigation to reach its target with high precision, as opposed to a ballistic missile., and radar technology that saved the United States an estimated $2 billion in weapons research[s]. The Washington Post later called him “one of CIA’s most valuable human assets in the Soviet Union”[s].
The CIA recruited Tolkachev using the SADRAT cycleThe CIA's six-step agent recruitment process: Spotting, Assessing, Developing, Recruiting, Agent handling, and Termination of a human intelligence source.: spotting potential sources, assessing their access and motivation, developing the relationship, recruiting formally, handling the agent operationally, and eventually terminating the relationship[s]. Russian intelligence services, drawing on over a century of continuous experience from the Cheka through the KGB to today’s SVR and GRU, still rely on fundamental principles: compartmentalizationA security principle that limits access to information by dividing it into separate, need-to-know segments to prevent unauthorized disclosure., deniability, patience, and psychological control of assets[s].
The Digital Surveillance Challenge
The 21st century brought challenges Cold War spymasters never imagined. China now operates an estimated 540 million surveillance cameras, with Beijing claiming total CCTV coverage of the capital[s]. In 2014, Russia adopted Executive Order 735 implementing biometric collection for all foreign visa applicants. In 2017, China began fingerprinting all foreigners at entry points[s]. These measures make alias travel extraordinarily difficult: biometric features are unique and hard to fake.
The 2010 Dubai assassination of Hamas commander Mahmoud Al-Mabhouh demonstrated operational vulnerability. Basic CCTV cameras across luxury hotels allowed investigators to track the Israeli team’s movements, disguise changes, and entry-exit patterns, completely unraveling the operation[s]. Biometrics now tie biological characteristics to specific identities across interconnected databases, with growing information sharing between allied nations limiting the old “one country, one alias” rule[s].
Digital footprints compound the challenge. Every device interaction generates traceable data. Cover identities now require years of backstopped social media presence. Facebook’s automated facial recognitionThe automated identification of individuals by analyzing facial features in images or video using AI algorithms. A match is an investigative lead, not proof of identity. might tag an operative’s cover profile, linking legend to true identity. The old paradigm of human intelligence, where a forged passport and solid legend carried an operative across borders, has fundamentally shifted.
Counter-Terror and the HUMINT Renaissance
Paradoxically, September 11, 2001 revealed that human intelligence remained essential when technical collection dominated. The War on Terror demanded renewed emphasis on infiltrating terrorist organizations[s]. According to former CIA Director John Brennan, the United States misjudged Saddam Hussein’s weapons capabilities in 2003 largely because it lacked human intelligence assets in Iraq[s].
Human intelligence may constitute only 10 to 20 percent of total intelligence inputs, with technical collection providing the remainder[s]. But human sources provide precisely what technology cannot: adversary intentions. Even in the drone age, targeting terrorist leadership requires human intelligence because these individuals minimize electronic communications and conceal themselves from imagery surveillance. Over-reliance on technical collection for drone targeting led to civilian deaths when imagery alone could not distinguish combatants from farmers[s].
Stuxnet: When Cyber Required Humans
The Stuxnet operation against Iran’s Natanz enrichment facility, discovered in 2010, became the defining example of cyber-physical fusion requiring human intelligence. The malware, reportedly developed jointly by the NSA, CIA, and Israel’s Unit 8200, was engineered to destroy gas centrifuges. But Natanz was air-gapped, physically isolated from the internet. Technology alone could not reach it.
Intelligence sources widely suggest Stuxnet entered via an infected USB flash drive, carried into the secure facility by an unsuspecting contractor or recruited insider[s]. The worm lay dormant until detecting specific Siemens controllers, then subtly altered centrifuge speeds while feeding false data to monitoring systems. Iranian engineers blamed equipment failure until the damage was irreversible. Approximately 1,000 of Iran’s 5,000 centrifuges were destroyed, delaying the nuclear program by an estimated two years[s].
Stuxnet proved the pattern: hostile actors use insiders or access agents to provide unique access impossible through cyber means alone[s]. As network defenses improve and air-gapping spreads, human intelligence becomes more critical for delivering digital payloads.
Russian Tradecraft Persists
Operation Ghost Stories, the FBI’s decade-long investigation culminating in June 2010 arrests, exposed ten SVR illegalsIntelligence operatives deployed abroad under false identities without diplomatic cover, operating as ordinary citizens while conducting espionage. operating across the United States[s]. These deep-cover operatives, some using stolen identities, established seemingly normal American lives while spotting and assessing potential recruitment targets. They never obtained classified documents, but they were playing the long game: developing sources in policymaking circles who might one day wield influence[s].
The approach echoed the Cambridge Five, where Soviet talent spotters recruited Cambridge University students including Kim Philby in the 1930s, then waited decades for them to penetrate British intelligence[s]. Russia still invests in long-term human intelligence operations because the payoff justifies decades of patience.
Russian tradecraft has adapted technologically while preserving fundamentals. Dead drops, what Russians call “tainiki,” remain standard operational practice: magnets concealed under bridges, hollowed-out stones, waterproof capsules buried in parks[s]. Agent signaling uses urban geography like specific telephone poles or mailboxes, the same method the KGB used with John Walker in the 1980s[s]. Communications now include encrypted messaging apps, laptop covert exchanges, and steganographyThe practice of hiding secret messages within ordinary text or images so the existence of the message itself is concealed. in digital images, but the underlying logic of compartmentalized, deniable contact persists.
Human Intelligence and Tomorrow’s Threats
In November 2021, MI6 Chief Richard Moore gave a rare public speech addressing the paradox facing modern intelligence. “To stay secret, we are going to have to become more open,” he said, explaining that technological progress in the next decade might equal the previous century[s]. But Moore emphasized the enduring necessity of human sources: “What we do, as a human intelligence agency, is essential, because at the end of the day, even in a digital world, critical decisions are made by real people”[s].
The United States now allocates $72.4 billion annually to its 18 intelligence agencies, employing approximately 100,000 people plus hundreds of thousands of contractors[s]. Technical collection absorbs most of this: satellites, signals intercepts, cyber operations. Yet human intelligence persists because no amount of technical prowess replaces a well-placed source with direct access to adversary intentions.
The spy trade survived by adapting. Cover identities now require years of digital backstopping. Biometric challenges force creative workarounds. Ubiquitous surveillance demands more sophisticated detection routes. But the fundamentals, recruiting human sources willing to betray secrets, remain unchanged since ancient times. The brush pass in a crowded market, the chalk mark on a lamppost, the hollow rock in a quiet park, the cultivated relationship yielding information no algorithm can see: these techniques endure because technology cannot replicate human motivation, trust, or judgment. Human intelligence survived the digital age by becoming inseparable from it.
