The transition to a new presidential administration has coincided with several thoughtful assessments on the future of U.S. intelligence that warn of an urgent need for reinvention in response to rapidly evolving circumstances and threats. For the most part, these have focused on the disruptive ramifications of new technologies — like artificial intelligence, machine learning, the internet-of-things, big data analytics, fifth-generation mobile technology, quantum computing, and synthetic biology — arguing that their emergence demands a fundamental rethinking of the country’s intelligence enterprise to exploit these developments and thwart their use against the United States. These assessments have largely been at the strategic level, touching on the full range of our complex Intelligence Community (IC), its various agencies, disciplines, and INT’s, including human intelligence (HUMINT).

This discussion of HUMINT often outlines clear challenges, such as threats posed by the increasingly pervasive use of tracking, or trackable, technologies that can be leveraged to form an exponentially more restrictive surveillance environment for those needing to operate clandestinely. Some observers point out, not without justification, that the culture of the U.S. HUMINT community has been historically resistant to adopting new technology, and to change in general. They argue that HUMINT operators, and HUMINT as a craft, will become irrelevant by failing to quickly embrace new technologies and increase the number of tech-savvy digital warriors less beholden to such “romantic” notions as believing they need to meet with their human agents in person. While these observations have merit, there is risk of their oversimplification to the detriment of the mission.

U.S. HUMINT operations could certainly benefit from an infusion of inventiveness, but the overwhelming emphasis placed on new technology leaves too much out, and arguably the most important parts. The reasoning of most recent assessments stressing technology is founded on the presumption of a “tech-driven future.” In fact, the Intelligence Community needs to cultivate not just cutting-edge technology, but human creativity and inventiveness writ large, because while mankind’s future may be technology-influenced, it is and will always be people-driven.

This is an important distinction. None other than Apple founder Steve Jobs famously conceded back in 2011 that the secret of Apple’s continuous disruptive success was not its technological prowess, but its merger of technology, liberal arts, and the humanities into the company’s DNA and its determination to tap not merely the talents of computer scientists and technologists, but “musicians, poets, artists, zoologists, and historians” as well. This suggests that a push by the IC to hire more STEM graduates is not likely to remedy its creativity and innovation deficit.

These days much is written about the impact of technology on HUMINT, but less on how human factors remain central to the application of technology to the intelligence craft beyond HUMINT. Regarding cyber threats and opportunities, for example, the IC places more emphasis on technical, hardware/software factors, such as zero-day exploits, firewalls, sniffers, scanners, and anti-virus programs, than on the “social engineering” aspects that are grounded in a sophisticated understanding of the humans who design and use these networks and the nuances of their behavior, sometimes referred to as “wetware.” As targeted networks become more technically hardened, or are detached from the world wide web altogether, it will also be more difficult to infiltrate tools, and exfiltrate the intelligence they collect, from the outside. It thus stands to reason that the requirement for cooperative-insider operations, i.e., hybrid HUMINT/cyber operations, will only grow.

Future technical operations will also remain at least partially dependent on various forms of human agency for their effective employment. Whereas some nodes of planned multimodal sensor arrays will be able to be used from stand-off distances, many intended to collect on the United States’ most menacing adversaries operating within denied areas will require closer proximity to their targets. Because of the hostile environments, achieving this proximity will likely require some form of HUMINT tradecraft.

Even aspects of open-source intelligence (OSINT) collection will occasionally require the employment of HUMINT-like practices. The availability and flow of information in authoritarian countries like Russia, China, Iran, and North Korea — the IC’s principal targets — is tightly controlled. Much information that would be open or commercially available in most other countries is off limits there. The independent investigative entity Bellingcat, for example, often held up as an example of what can be divined using open source methods, candidly admits that key elements of the data it acquires comes from leaked or pilfered datasets purchased on the black market in denied area countries using obscured payment mechanisms, and that the purveyors of this information can be, and have been, subjected to criminal prosecution in their respective countries. Thus, even open-source collection can at times involve shades of gray and the necessity for confidentiality and source protection tradecraft.

There is truth to the assertion that there is institutional resistance to change and the adoption of new technologies within the HUMINT community, but ascribing this to romanticism or some atavistic devotion to outdated traditions is an unfair caricature. The real explanation is more complex and thoughtful than is given credit for. HUMINT professionals would counter that those who advocate a broad transition to “fully digitized” recruitment and handling of human sources underappreciate a key element of HUMINT: the humans. In the authoritarian regimes that are the United States’ highest priority targets, individuals who agree to clandestine cooperation with U.S. intelligence live under a level of emotional strain that is difficult to overstate. They do so at the risk of their liberty, their lives, and sometimes those of their families. Their motivations are powerful, complex, diverse, and often dynamic. HUMINT professionals feel this most directly, and for those who have suffered the loss of an agent, the experience is searing and deeply felt at a personal level. This is the real source of the conservatism vis-a-vis technologies that are perceived to be untested. Not romanticism or a devotion to tradition, but a visceral devotion to the safety of human sources.

In the real world, some agents are comfortable with impersonal communication arrangements not requiring face-to-face meetings. In fact, some demand it. Others have a distrust of advanced technology, lack confidence in their competence to use it safely, or simply have a powerful need for human contact with their handler, who is probably the only other person they can talk to about what is likely the deepest secret of their lives. Many factors go into the calculation of the means of clandestine communication to be used with an agent. Suggesting a broad shift to virtual agent recruitment and handling based on the latest digital technology risks replacing one dogma with another.

Furthermore, reliance on new technologies to avoid personal meetings comes with its own set of dangers. Osama bin Laden escaped justice for almost a decade after 9/11 not because of his savvy use of modern technology, but because of his disciplined avoidance of it. In some clandestine HUMINT scenarios, the best solution may not be the adoption of a new technology, but an asymmetrical response that dusts off and modernizes older technologies and techniques. This is not to advocate a revival in the use of carrier pigeons and No.2 pencils, but the Intelligence Community would be wise to recognize that creativity and innovation are not solely the domain of shiny new things, and saying this does not make one a Luddite. Human sources can of course be recruited and run virtually, and we should look to expand this practice to the degree practicable, but it is unrealistic to expect that this could or should ever become the backbone of U.S. clandestine HUMINT operations.

The above is not to suggest that there is not serious room for greater imagination, innovation, and flexibility in our HUMINT practices. Ironically, the most surprising shortcoming among HUMINT operators is not their reluctance to adopt new technology, but the lack of a systematic professional grounding in advanced concepts related to human behavior, motivation, and decision-making. This is nothing short of shocking for a cadre that is in the people business–by definition. It is likely not an exaggeration to assert that the sales staffs of most large-to-mid-sized U.S. companies are more conversant than the average HUMINT case officer with the writings and concepts of such key psychology and behavioral science authorities as Viktor Frankl, Abraham Maslow, Daniel Kahneman, Richard Thaler, or Robert Cialdini. If the business of HUMINT is people, and the goal is virtuosity, this must change.

A certain rigidity has also crept into the CIA’s contemporary HUMINT doctrine regarding what constitutes HUMINT and how it should be pursued. The gold standard has always been the in-place agent. It is the most difficult to accomplish, but usually also the most valuable given the potential for a continuous stream of fresh information and the ability to task. It is HUMINT’s equivalent of cyber’s Advanced Persistent Threat (APT). Over time, however, the preference for in-place sources has evolved into a near exclusive requirement. The IC should not lose sight of the fact that historically blockbuster intelligence has also come from defectors, such as the reporting that blew the lid off Soviet technical efforts to penetrate the US Embassy in Moscow and the massive and treaty-violating Soviet/Russian biological warfare program. Critical, and sometimes chilling, reporting has also come from debriefing emigres, yet these variations of the HUMINT craft, once core pursuits, have fallen out of favor in the current orthodoxy despite their potential to fill critical intelligence gaps in an environment where it is becoming more challenging to run in-place sources in denied areas. Perhaps this is worth a fresh look in the spirit of reinvention.

As with the rest of the IC, there is indeed an urgent need for renewed imagination in the practice of the HUMINT craft. The lopsided emphasis on technology in recent assessments, however, risks an under-appreciation of the even greater need to foster an expansion of human creativity and inventiveness overall, in which new technology will play a significant but hardly exclusive role. Organizational culture should not be reduced to being thought of merely as an obstacle or enabler to the adoption of new technology; it is the thing itself, for it is the human dimension, not technology, that will be the key determinant of ultimate success in reimagining intelligence.